
# data analysis and wrangling
import pandas as pd
import numpy as np
import random as rnd

# visualization
import seaborn as sns
import matplotlib.pyplot as plt
%matplotlib inline
pd.set_option('display.max_columns', 50)
sns.set(style='darkgrid')
sns.set_palette('husl', 8)
plt.rcParams['font.family'] = 'Times New Roman'

# machine learning
from sklearn.linear_model import LogisticRegression
from sklearn.svm import SVC, LinearSVC
from sklearn.ensemble import RandomForestClassifier
from sklearn.neighbors import KNeighborsClassifier
from sklearn.naive_bayes import GaussianNB
from sklearn.linear_model import Perceptron
from sklearn.linear_model import SGDClassifier
from sklearn.tree import DecisionTreeClassifier


filepath = r'~/Desktop/Desktop/Prasmul/Semester 6/Applied Data Science for Business/Project Group Dataset/'
filename = r'cctx_train.xlsx'

data = pd.read_excel(filepath + filename)
data.head()


data.info(memory_usage='deep')

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10500 entries, 0 to 10499
Data columns (total 21 columns):
 #   Column                         Non-Null Count  Dtype         
---  ------                         --------------  -----         
 0   Transaction_ID                 10500 non-null  object        
 1   Transaction_Flag               10500 non-null  bool          
 2   Transaction_Date               10500 non-null  datetime64[ns]
 3   Transaction_Type               10500 non-null  object        
 4   Transaction_Amount             10500 non-null  float64       
 5   Bank_ID                        10500 non-null  int64         
 6   CC_ID                          10500 non-null  object        
 7   Card_Type                      10500 non-null  object        
 8   Card_Holder                    10500 non-null  int64         
 9   Channel_ID                     10500 non-null  int64         
 10  Merchant_ID                    10500 non-null  object        
 11  Country_ID                     10500 non-null  object        
 12  City_ID                        10500 non-null  object        
 13  EDC_Type                       10500 non-null  object        
 14  EDC_Location                   10500 non-null  object        
 15  EDC_Owner                      10500 non-null  object        
 16  Average_Transaction_Amount     10484 non-null  float64       
 17  Maximum_Transaction_Amount     10484 non-null  float64       
 18  Minimum_Transaction_Amount     10484 non-null  float64       
 19  Average_Transaction_Frequency  10484 non-null  float64       
 20  Fraud_Status                   10500 non-null  int64         
dtypes: bool(1), datetime64[ns](1), float64(5), int64(4), object(10)
memory usage: 7.2 MB


def null_counts(df, style=True):
    nulls = df.isna().sum().rename_axis('Columns').reset_index(name='Count')
    nulls['Percentage'] = nulls['Count'] / len(df)
    nulls = nulls.loc[nulls['Count'] > 0]
    nulls.sort_values(by='Count', ascending=False, inplace=True)
    nulls.reset_index(drop=True, inplace=True)
    
    if style:
        nulls = nulls.style.format({'Count': '{:,}', 'Percentage': '{:.2%}'}).hide_index()
        
    return nulls


null_counts(data)


data.dropna()


data.describe()


data.describe(include=['O'])


data = data.drop(['Transaction_ID','Transaction_Flag','Country_ID','Bank_ID','CC_ID','EDC_Location'], axis=1)
data.head()


def value_counts(df, col, style=True):
    table = df[col].value_counts().rename_axis('Value').reset_index(name='Count')
    table['Percentage'] = table['Count'] / table['Count'].sum(axis=0)
    
    if style:
        table = table.style.format({'Count': '{:,}', 'Percentage': '{:.2%}'}).hide_index()
        
    return table


value_counts(data, 'Fraud_Status')


labels=['Not Fraud','Fraud']
colors=['green', 'red']
data['Fraud_Status'].value_counts().sort_values(ascending=False).plot.pie(autopct='%1.1f%%', labels=labels, colors=colors, radius=1.5)

<AxesSubplot:ylabel='Fraud_Status'>


value_counts(data, 'Card_Type')


data['Card_Type'].replace({'CC11': 0, 'CC08': 1, 'CC09':2, 'CC00':3, 'CC10': 4}, inplace=True)
data.head()


unused = data['Card_Type'].loc[(data['Card_Type']!= 0)&(data['Card_Type']!=1)& (data['Card_Type']!=2) & (data['Card_Type']!=3) &(data['Card_Type']!=4)]
unused

4        CC12
11       CC04
19       CC12
23       CC12
25       CC01
         ... 
10436    CC02
10445    CC01
10462    CC04
10465    CC01
10487    CC02
Name: Card_Type, Length: 812, dtype: object


data.replace(unused.values, 5, inplace=True)
data


value_counts(data, 'Transaction_Type')


data['Transaction_Type'].replace({'T01': 0, 'T15': 1, 'T02':2, 'T08':3}, inplace=True)
data.head()


unused = data['Transaction_Type'].loc[(data['Transaction_Type']!= 0)&(data['Transaction_Type']!=1)& (data['Transaction_Type']!=2) & (data['Transaction_Type']!=3)]
unused

4        T10
5        T03
6        T13
8        T06
9        T06
        ... 
10493    T06
10495    T13
10496    T13
10497    T13
10498    T13
Name: Transaction_Type, Length: 2781, dtype: object


data.replace(unused.values, 4, inplace=True)
data


value_counts(data, 'Card_Holder')


data['Card_Holder'].replace({1: 0, 2: 1}, inplace=True)
data.head()


value_counts(data, 'EDC_Type')


data['EDC_Type'].replace({'EDC0002':0}, inplace=True)
data.head()


unused = data['EDC_Type'].loc[data['EDC_Type']!= 0]
unused

0        EDC0885
1        EDC0565
2        EDC4639
3        EDC3918
4        EDC2863
          ...   
10495    EDC0741
10496    EDC0797
10497    EDC0587
10498    EDC0626
10499    EDC4477
Name: EDC_Type, Length: 9793, dtype: object


data.replace(unused.values, 1, inplace=True)
data


value_counts(data, 'Channel_ID')


data['Channel_ID'].replace({1:1}, inplace=True)
data.head()


unused = data['Channel_ID'].loc[data['Channel_ID']!= 1]
unused

8        5
9        5
13       5
21       5
29       5
        ..
10471    5
10475    2
10492    5
10493    5
10499    2
Name: Channel_ID, Length: 2175, dtype: int64


data.replace(unused.values, 0, inplace=True)
data


value_counts(data, 'Merchant_ID')


data['Merchant_ID'].replace({'M0001':1}, inplace=True)
data.head()


unused = data['Merchant_ID'].loc[data['Merchant_ID']!= 1]
data.replace(unused.values, 0, inplace=True)
data


value_counts(data, 'City_ID')


data['City_ID'].replace({'CTY06-023':0}, inplace=True)
data.head()


unused = data['City_ID'].loc[data['City_ID']!= 0]
data.replace(unused.values, 1, inplace=True)
data


value_counts(data, 'EDC_Owner')


data['EDC_Owner'].replace({'OEDC0377':1}, inplace=True)
data.head()


unused = data['EDC_Owner'].loc[data['EDC_Owner']!= 1]
data.replace(unused.values, 0, inplace=True)
data


data['Month'] = data['Transaction_Date'].dt.month_name()
data['Day'] = data['Transaction_Date'].dt.day_name()
data.head()


data['Day'].replace({'Sunday': 0, 'Monday': 1, 'Tuesday':2, 'Wednesday':3, 'Thursday': 4, 'Friday': 5, 'Saturday': 6, 'Sunday': 7}, inplace=True)
data.head()


data['Month'].replace({'January': 0, 'February': 1, 'March':2, 'April':3, 'May': 4, 'June': 5, 'July': 6, 'August': 7,'September':8,'October':9,'November':10,'December':11}, inplace=True)
data.head()


data.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 10500 entries, 0 to 10499
Data columns (total 17 columns):
 #   Column                         Non-Null Count  Dtype         
---  ------                         --------------  -----         
 0   Transaction_Date               10500 non-null  datetime64[ns]
 1   Transaction_Type               10500 non-null  int64         
 2   Transaction_Amount             10500 non-null  float64       
 3   Card_Type                      10500 non-null  int64         
 4   Card_Holder                    10500 non-null  int64         
 5   Channel_ID                     10500 non-null  int64         
 6   Merchant_ID                    10500 non-null  int64         
 7   City_ID                        10500 non-null  int64         
 8   EDC_Type                       10500 non-null  int64         
 9   EDC_Owner                      10500 non-null  int64         
 10  Average_Transaction_Amount     10484 non-null  float64       
 11  Maximum_Transaction_Amount     10484 non-null  float64       
 12  Minimum_Transaction_Amount     10484 non-null  float64       
 13  Average_Transaction_Frequency  10484 non-null  float64       
 14  Fraud_Status                   10500 non-null  int64         
 15  Month                          10500 non-null  int64         
 16  Day                            10500 non-null  int64         
dtypes: datetime64[ns](1), float64(5), int64(11)
memory usage: 1.4 MB


data=data.loc[data['Transaction_Amount']<data['Maximum_Transaction_Amount']]
data.head()


data=data.loc[data['Transaction_Amount']>data['Minimum_Transaction_Amount']]
data.head()


response = 'Fraud_Status'
unused_glm_vars = ['Transaction_Type','EDC_Type', 'EDC_Owner','Fraud_Status','Transaction_Type']
glm_vars = list(set(data.columns) - set(unused_glm_vars))
glm_vars

['Channel_ID',
 'Day',
 'Merchant_ID',
 'Card_Type',
 'Maximum_Transaction_Amount',
 'Average_Transaction_Frequency',
 'City_ID',
 'Transaction_Amount',
 'Minimum_Transaction_Amount',
 'Month',
 'Average_Transaction_Amount',
 'Transaction_Date',
 'Card_Holder']


def target_boxplot(data, variable, target='Target', plotsize=(8, 8), colors=['green', 'red'], labels=['No', 'Yes'], limits=None):
    fig, ax = plt.subplots(figsize=plotsize)
    
    bp = sns.boxplot(x=target, y=variable, data=data, showmeans=True, meanline=True, 
                     palette=colors, width=0.3,
                     flierprops=dict(marker='o', markerfacecolor='red', markeredgecolor='white', alpha=0.5),
                     meanprops=dict(linestyle='-', linewidth=2, color='orange'),
                     medianprops=dict(linestyle='-', linewidth=2, color='white'), 
                     boxprops=dict(linewidth=0, alpha=0.7))
    
    bp.set(xticklabels=labels, ylim=limits)
    
    plt.show()


def bar_plot(pivot_table, color=['green', 'red'], size=(15,15), d=0.05):
    ax = pivot_table.plot(kind='barh', figsize=size, fontsize=12, color=color, alpha=0.7)
    for val in ax.patches:
        width = val.get_width()
        height = val.get_height()
        x, y = val.get_xy() 
        ax.annotate(f'{height:.0%}', (x + height/2, y + width*1.02), ha='center')

    ax.set_ylabel('')
    ax.set_xticks([])
    plt.show()


colors_list = ['#5cb85c','#5bc0de','#d9534f']

# Normalize result
result_pct = result.div(result.sum(1), axis=0)

ax = result_pct.plot(kind='bar',figsize=(15,4),width = 0.8,color = colors_list,edgecolor=None)
plt.legend(labels=result.columns,fontsize= 14)
plt.title("Percentage of Respondents' Interest in Data Science Areas",fontsize= 16)

plt.xticks(fontsize=14)
for spine in plt.gca().spines.values():
    spine.set_visible(False)
plt.yticks([])

# Add this loop to add the annotations


for var in ['Transaction_Amount', 'Average_Transaction_Frequency', 'Maximum_Transaction_Amount', 'Average_Transaction_Amount','Minimum_Transaction_Amount']:
    plt.figure()
    target_boxplot(data, var, target=response, labels=['Not Fraud', 'Fraud'])

<Figure size 432x288 with 0 Axes>

<Figure size 432x288 with 0 Axes>

<Figure size 432x288 with 0 Axes>

<Figure size 432x288 with 0 Axes>

<Figure size 432x288 with 0 Axes>


pvt = pd.crosstab(index=data['Month'], columns=data['Fraud_Status'])
pvt


bar_plot(pvt, d=0.05)

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
/Applications/anaconda3/lib/python3.8/site-packages/IPython/core/formatters.py in __call__(self, obj)
    339                 pass
    340             else:
--> 341                 return printer(obj)
    342             # Finally look for special method names
    343             method = get_real_method(obj, self.print_method)

/Applications/anaconda3/lib/python3.8/site-packages/IPython/core/pylabtools.py in <lambda>(fig)
    246 
    247     if 'png' in formats:
--> 248         png_formatter.for_type(Figure, lambda fig: print_figure(fig, 'png', **kwargs))
    249     if 'retina' in formats or 'png2x' in formats:
    250         png_formatter.for_type(Figure, lambda fig: retina_figure(fig, **kwargs))

/Applications/anaconda3/lib/python3.8/site-packages/IPython/core/pylabtools.py in print_figure(fig, fmt, bbox_inches, **kwargs)
    130         FigureCanvasBase(fig)
    131 
--> 132     fig.canvas.print_figure(bytes_io, **kw)
    133     data = bytes_io.getvalue()
    134     if fmt == 'svg':

/Applications/anaconda3/lib/python3.8/site-packages/matplotlib/backend_bases.py in print_figure(self, filename, dpi, facecolor, edgecolor, orientation, format, bbox_inches, pad_inches, bbox_extra_artists, backend, **kwargs)
   2208 
   2209             try:
-> 2210                 result = print_method(
   2211                     filename,
   2212                     dpi=dpi,

/Applications/anaconda3/lib/python3.8/site-packages/matplotlib/backend_bases.py in wrapper(*args, **kwargs)
   1637             kwargs.pop(arg)
   1638 
-> 1639         return func(*args, **kwargs)
   1640 
   1641     return wrapper

/Applications/anaconda3/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py in print_png(self, filename_or_obj, metadata, pil_kwargs, *args)
    507             *metadata*, including the default 'Software' key.
    508         """
--> 509         FigureCanvasAgg.draw(self)
    510         mpl.image.imsave(
    511             filename_or_obj, self.buffer_rgba(), format="png", origin="upper",

/Applications/anaconda3/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py in draw(self)
    400     def draw(self):
    401         # docstring inherited
--> 402         self.renderer = self.get_renderer(cleared=True)
    403         # Acquire a lock on the shared font cache.
    404         with RendererAgg.lock, \

/Applications/anaconda3/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py in get_renderer(self, cleared)
    416                           and getattr(self, "_lastKey", None) == key)
    417         if not reuse_renderer:
--> 418             self.renderer = RendererAgg(w, h, self.figure.dpi)
    419             self._lastKey = key
    420         elif cleared:

/Applications/anaconda3/lib/python3.8/site-packages/matplotlib/backends/backend_agg.py in __init__(self, width, height, dpi)
     94         self.width = width
     95         self.height = height
---> 96         self._renderer = _RendererAgg(int(width), int(height), dpi)
     97         self._filter_renderers = []
     98 

ValueError: Image size of 874x89094 pixels is too large. It must be less than 2^16 in each direction.

<Figure size 1080x1080 with 1 Axes>


data['Transaction_Type'].value_counts()

0    7417
1    2316
Name: Transaction_Type, dtype: int64


cardtype=data[["Card_Type", 'Fraud_Status' ]].groupby(["Card_Type"], as_index=False).mean().sort_values(by='Fraud_Status', ascending=False)
fcardtype=cardtype.head()
fcardtype


sns.set_theme(style="whitegrid")
ax = sns.barplot(x="Card_Type", y="Fraud_Status", data=fcardtype)


cardholder=data[["Card_Holder", 'Fraud_Status' ]].groupby(["Card_Holder"], as_index=False).mean().sort_values(by='Fraud_Status', ascending=False)
cardholder


sns.set_theme(style="whitegrid")
ax = sns.barplot(x="Card_Holder", y="Fraud_Status", data=cardholder)


data.groupby('Card_Type')['Fraud_Status'].value_counts().unstack().stack(dropna=False).reset_index(name="Count").set_index(['Card_Type', 'Fraud_Status'])


cardholder=data.groupby('Card_Holder')['Fraud_Status'].value_counts().unstack().stack(dropna=False).reset_index(name="Count").set_index(['Card_Holder', 'Fraud_Status'])
cardholder


cardholder.unstack().plot(kind='bar', stacked=True)

<AxesSubplot:xlabel='Card_Holder'>


def correlation_plot(df, plotsize=(15,15), corr_type='spearman', number_format='.1%', fontsize=11):
    corrs = df.corr(method=corr_type)
    triangle = np.triu(corrs)

    fig, ax = plt.subplots(figsize=plotsize)
    ax = sns.heatmap(corrs, mask=triangle, annot=True, annot_kws={'size': fontsize}, fmt=number_format, cmap='RdYlGn_r', 
                     cbar_kws={'shrink': 0.8}, vmin=-1, center=0, vmax=1, square=True, linewidths=0.1, linecolor='white')
    plt.show()


correlation_plot(data, fontsize=14)


import ppscore as pps
from scipy.stats import pearsonr
from scipy.stats import spearmanr
from itertools import combinations

def pps_correlations(df, variables, corr_type='pearson', limit=0.8, style=True):
    
    vars1, vars2 = [], []
    corrs, pvals = [], []
    ppscores1, ppscores2 = [], []
    
    for var1, var2 in combinations(variables, 2):
        if corr_type == 'pearson':
            corr, pval = pearsonr(df[var1], df[var2])
        elif corr_type == 'spearman':
            corr, pval = spearmanr(df[var1], df[var2])
        else:
            print('Unknown correlation method. Please input either \'pearson\' or \'spearman\' in argument corr_type.')
            break
        
        vars1.append(var1)
        vars2.append(var2)
        corrs.append(corr)
        pvals.append(pval)
        
    ppcorr = pd.DataFrame({'Variable 1': vars1, 'Variable 2': vars2, 'Correlation': corrs, 'P-Value': pvals})
    cols = list(ppcorr.columns)
    
    ppcorr['Absolute'] = ppcorr['Correlation'].abs()
    ppcorr.sort_values(by='Absolute', ascending=False, inplace=True)
    ppcorr.reset_index(drop=True, inplace=True)
    ppcorr = ppcorr.loc[ppcorr['Absolute'] >= limit, cols].copy()
    
    vars1 = ppcorr['Variable 1'].tolist()
    vars2 = ppcorr['Variable 2'].tolist()
    
    for var1, var2 in zip(vars1, vars2):
        ppscores1.append(pps.score(df, var1, var2)['ppscore'])
        ppscores2.append(pps.score(df, var2, var1)['ppscore'])
        
    ppcorr['PPS 1'] = ppscores1
    ppcorr['PPS 2'] = ppscores2
    cols.insert(2, 'PPS 2')
    cols.insert(2, 'PPS 1')
    ppcorr = ppcorr[cols]
    ppcorr.index += 1
    
    if style:
        formats = {'PPS 1': '{:.2%}', 'PPS 2': '{:.2%}', 'Correlation': '{:.2%}', 'P-Value': '{:.2%}'}
        ppcorr = ppcorr.style.format(formats)
    
    return ppcorr


from statsmodels.tools.tools import add_constant
from statsmodels.stats.outliers_influence import variance_inflation_factor

def vif(data, features, style=True):
    # Add constant to the GLM variables
    X = add_constant(data[features])
    vif_values = [variance_inflation_factor(X.values, i) for i in range(1, X.shape[1])]
    vif_df = pd.DataFrame({'Variable': features, 'VIF': vif_values})
    vif_df = vif_df.sort_values('VIF', ascending=False).reset_index(drop=True)
    vif_df.index += 1

    if style:
        vif_df = vif_df.style.format({'VIF': '{:,.2f}'})
        
    return vif_df


vif(data, glm_vars)

/Applications/anaconda3/lib/python3.8/site-packages/statsmodels/stats/outliers_influence.py:193: RuntimeWarning: divide by zero encountered in double_scalars
  vif = 1. / (1. - r_squared_i)


pps_correlations(data, glm_vars, limit=0)


import h2o
h2o.init()

from h2o.automl import H2OAutoML

Checking whether there is an H2O instance running at http://localhost:54321 . connected.


def timer(start_time, header='\nProgram Schedule'):
    end_time = pd.to_datetime('now') + pd.Timedelta('07:00:00')
    print(header)
    print(start_time.strftime('Start   : %Y-%m-%d %I:%M:%S %p'))
    print(end_time.strftime('Finish  : %Y-%m-%d %I:%M:%S %p'))
    run_time = str(end_time - start_time)
    idx1 = run_time.find(' ', 3) + 1
    idx2 = run_time.find('.')
    print('Runtime :', run_time[idx1 : idx2], '\n')


from sklearn.model_selection import train_test_split

start_time = pd.to_datetime('now') + pd.Timedelta('07:00:00')
trains=[]
tests=[]

for i in range(0, 30):
    
    train, test = train_test_split(data, test_size=0.2, stratify=data['Fraud_Status'])
    train = h2o.H2OFrame(train)
    test = h2o.H2OFrame(test)
    train['Fraud_Status'] = train['Fraud_Status'].asfactor()
    test['Fraud_Status'] = test['Fraud_Status'].asfactor()
    train['Transaction_Type'] = train['Transaction_Type'].asfactor()
    test['Transaction_Type'] = test['Transaction_Type'].asfactor()
    train['Card_Type'] = train['Card_Type'].asfactor()
    test['Card_Type'] = test['Card_Type'].asfactor()
    trains.append(train)
    tests.append(test)

timer(start_time, header='\nConverting DataFrame to H2O Frame')

Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%
Parse progress: |█████████████████████████████████████████████████████████| 100%

Converting DataFrame to H2O Frame
Start   : 2021-07-15 08:09:59 PM
Finish  : 2021-07-15 08:10:27 PM
Runtime : 00:00:27


start_time = pd.to_datetime('now') + pd.Timedelta('07:00:00')
models=[]
for train, test in zip(trains, tests):
    GLM = H2OAutoML(max_runtime_secs=300, nfolds=30, include_algos=['GLM'])
    GLM.train(x=glm_vars, y='Fraud_Status', training_frame=train, validation_frame=test)
    models.append(GLM.leader)
timer(start_time, header='\nModeling')

AutoML progress: |
20:11:08.884: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:11:21.258: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:11:31.783: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:11:40.639: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:11:51.835: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:12:03.146: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:12:13.541: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:12:25.781: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:12:35.780: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:12:49.225: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:12:58.494: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:13:10.62: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:13:19.401: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:13:31.994: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:13:42.919: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:13:54.756: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:14:06.719: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:14:19.528: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:14:30.869: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:14:43.292: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:14:54.424: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:15:05.296: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:15:17.961: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:15:33.585: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:15:48.315: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:16:01.238: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:16:11.651: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:16:23.833: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:16:34.61: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:16:47.766: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%

Modeling
Start   : 2021-07-15 08:11:08 PM
Finish  : 2021-07-15 08:16:59 PM
Runtime : 00:05:50


models

Model Details
=============
H2OGeneralizedLinearEstimator :  Generalized Linear Modeling
Model Key:  GLM_1_AutoML_20210715_201108


GLM Model: summary


ModelMetricsBinomialGLM: glm
** Reported on train data. **

MSE: 0.04752236156796884
RMSE: 0.21799624209597934
LogLoss: 0.1840262580204793
Null degrees of freedom: 7785
Residual degrees of freedom: 7771
Null deviance: 3538.9592337938643
Residual deviance: 2865.6568898949035
AIC: 2895.6568898949035
AUC: 0.7794022408158785
AUCPR: 0.31251018779415657
Gini: 0.558804481631757

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.29512062976335124:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  6.01 %


ModelMetricsBinomialGLM: glm
** Reported on validation data. **

MSE: 0.05027232283187711
RMSE: 0.22421490323320864
LogLoss: 0.19448948179990513
Null degrees of freedom: 1946
Residual degrees of freedom: 1932
Null deviance: 884.8017986311489
Residual deviance: 757.3420421288304
AIC: 787.3420421288304
AUC: 0.7457078137406006
AUCPR: 0.22155303320166622
Gini: 0.49141562748120116

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.10879492515232687:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  5.79 %


ModelMetricsBinomialGLM: glm
** Reported on cross-validation data. **

MSE: 0.047653128252676175
RMSE: 0.21829596481079575
LogLoss: 0.18568864879336736
Null degrees of freedom: 7785
Residual degrees of freedom: 7771
Null deviance: 3541.0611592595515
Residual deviance: 2891.5436390103164
AIC: 2921.5436390103164
AUC: 0.7634122220587101
AUCPR: 0.3034137782994518
Gini: 0.5268244441174201

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.11952626883590241:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  5.99 %


import statistics
totalF1 = []
for i in models:
    F1s=i.F1(valid=True)[0][1]
    totalF1.append(F1s)
a=statistics.mean(totalF1)
a

0.41995830294642694


totalthreshold = []
for i in models:
    thresholds=i.F1(valid=True)[0][0]
    totalthreshold.append(thresholds)
b=statistics.mean(totalthreshold)
b

0.19787958062573607


totalprecision = []
for i in models:
    precisions=i.precision(valid=True)[0][1]
    totalprecision.append(precisions)
c=statistics.mean(totalprecision)
c

0.8318639291465378


totalrecall = []
for i in models:
    recalls=i.recall(valid=True)[0][1]
    totalrecall.append(recalls)
d=statistics.mean(totalrecall)
d

1.0


totalaccuracy = []
for i in models:
    accuracies=i.accuracy(valid=True)[0][1]
    totalaccuracy.append(accuracies)
e=statistics.mean(totalaccuracy)
e

0.9417907892484164


f1 = a
threshold = b
precision = c
sensitivity = d
accuracy = e
metrics = {'Threshold': threshold, 'F1': f1, 'Precision': precision, 'Sensitivity': sensitivity, 'Accuracy': accuracy}
glm_performance = pd.DataFrame(metrics.values(), columns=['Value'], index=metrics.keys())
glm_performance


base_model = GLM.leader
base_model

Model Details
=============
H2OGeneralizedLinearEstimator :  Generalized Linear Modeling
Model Key:  GLM_1_AutoML_20210715_201647


GLM Model: summary


ModelMetricsBinomialGLM: glm
** Reported on train data. **

MSE: 0.04813232537278134
RMSE: 0.2193908051235998
LogLoss: 0.18637933900220385
Null degrees of freedom: 7785
Residual degrees of freedom: 7771
Null deviance: 3538.959233793846
Residual deviance: 2902.299066942318
AIC: 2932.299066942318
AUC: 0.7707934772706568
AUCPR: 0.289212154938732
Gini: 0.5415869545413137

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.14313113283612258:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  6.01 %


ModelMetricsBinomialGLM: glm
** Reported on validation data. **

MSE: 0.04782450600267793
RMSE: 0.21868814783311402
LogLoss: 0.18447730336658244
Null degrees of freedom: 1946
Residual degrees of freedom: 1932
Null deviance: 884.8017986311447
Residual deviance: 718.3546193094721
AIC: 748.3546193094721
AUC: 0.7905539208817898
AUCPR: 0.29432069272826045
Gini: 0.5811078417635795

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.21146723168373893:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  5.85 %


ModelMetricsBinomialGLM: glm
** Reported on cross-validation data. **

MSE: 0.048314917844115436
RMSE: 0.21980654640868966
LogLoss: 0.1874930722653597
Null degrees of freedom: 7785
Residual degrees of freedom: 7771
Null deviance: 3540.536621270318
Residual deviance: 2919.642121316181
AIC: 2949.642121316181
AUC: 0.7614112433222846
AUCPR: 0.2789306901875544
Gini: 0.5228224866445692

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.2500729007254136:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  5.99 %


base_model.std_coef_plot(num_of_features=10)


def glm_coefficients(glm):
    coefs = glm._model_json['output']['coefficients_table'].as_data_frame()
    coefs['sign'] = np.where(coefs['standardized_coefficients'] < 0, 'Negative', 'Positive')
    coefs['standardized_absolutes'] = coefs['standardized_coefficients'].abs()
    coefs.sort_values(by='standardized_absolutes', ascending=False, inplace=True)
    coefs.reset_index(drop=True, inplace=True)
    
    return coefs


coefs_base_model = glm_coefficients(base_model)
coefs_base_model.head(20)


def variable_importances(model, limit=0.95):
    varimp = model.varimp(use_pandas=True)
    varimp['cumulative'] = varimp['percentage'].cumsum()
    varimp = varimp[varimp['cumulative'] <= limit].copy()
    
    return varimp


varimp_base_model = variable_importances(base_model, limit=0.99)
varimp_base_model


start_time = pd.to_datetime('now') + pd.Timedelta('07:00:00')
models=[]
for train, test in zip(trains, tests):
    GBM = H2OAutoML(max_runtime_secs=300, nfolds=30, include_algos=['DRF'])
    GBM.train(x=glm_vars, y='Fraud_Status', training_frame=train, validation_frame=test)
    models.append(GBM.leader)
timer(start_time, header='\nModeling')

AutoML progress: |█
20:46:37.669: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

███████████████████████████████████████████████████████| 100%
AutoML progress: |
20:47:39.28: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:48:43.327: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:49:53.725: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:50:57.432: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:52:03.778: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:53:14.246: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:54:30.561: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:55:38.595: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:56:48.853: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:57:57.741: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
20:59:07.160: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:00:15.552: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:01:24.31: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:02:10.351: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:02:57.506: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:03:49.894: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:04:47.24: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:05:48.354: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:06:51.427: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:07:57.575: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:09:14.671: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:10:35.143: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

███████████████████████████████████████████Failed polling AutoML progress log: Unexpected HTTP error: ('Connection aborted.', BadStatusLine('GET /3/Frames/AutoML_20210715_211035138_eventlog?row_count=10&row_offset=0&column_count=-1&full_column_count=-1&column_offset=0 HTTP/1.1\r\n'))
█████████████| 100%
AutoML progress: |
21:11:52.575: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:13:05.732: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:14:36.888: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:15:55.175: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:17:04.617: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:18:23.611: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%
AutoML progress: |
21:19:45.791: User specified a validation frame with cross-validation still enabled. Please note that the models will still be validated using cross-validation only, the validation frame will be used to provide purely informative validation metrics on the trained models.

████████████████████████████████████████████████████████| 100%

Modeling
Start   : 2021-07-15 08:46:37 PM
Finish  : 2021-07-15 09:20:57 PM
Runtime : 00:34:20


gbm = GLM.leader
gbm

Model Details
=============
H2OGeneralizedLinearEstimator :  Generalized Linear Modeling
Model Key:  GLM_1_AutoML_20210715_201647


GLM Model: summary


ModelMetricsBinomialGLM: glm
** Reported on train data. **

MSE: 0.04813232537278134
RMSE: 0.2193908051235998
LogLoss: 0.18637933900220385
Null degrees of freedom: 7785
Residual degrees of freedom: 7771
Null deviance: 3538.959233793846
Residual deviance: 2902.299066942318
AIC: 2932.299066942318
AUC: 0.7707934772706568
AUCPR: 0.289212154938732
Gini: 0.5415869545413137

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.14313113283612258:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  6.01 %


ModelMetricsBinomialGLM: glm
** Reported on validation data. **

MSE: 0.04782450600267793
RMSE: 0.21868814783311402
LogLoss: 0.18447730336658244
Null degrees of freedom: 1946
Residual degrees of freedom: 1932
Null deviance: 884.8017986311447
Residual deviance: 718.3546193094721
AIC: 748.3546193094721
AUC: 0.7905539208817898
AUCPR: 0.29432069272826045
Gini: 0.5811078417635795

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.21146723168373893:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  5.85 %


ModelMetricsBinomialGLM: glm
** Reported on cross-validation data. **

MSE: 0.048314917844115436
RMSE: 0.21980654640868966
LogLoss: 0.1874930722653597
Null degrees of freedom: 7785
Residual degrees of freedom: 7771
Null deviance: 3540.536621270318
Residual deviance: 2919.642121316181
AIC: 2949.642121316181
AUC: 0.7614112433222846
AUCPR: 0.2789306901875544
Gini: 0.5228224866445692

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.2500729007254136:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  5.99 %


totalF1 = []
for i in models:
    F1s=i.F1(valid=True)[0][1]
    totalF1.append(F1s)
a=statistics.mean(totalF1)
a

0.42137432219041765


totalthreshold = []
for i in models:
    thresholds=i.F1(valid=True)[0][0]
    totalthreshold.append(thresholds)
b=statistics.mean(totalthreshold)
b

0.24674137793754738


totalprecision = []
for i in models:
    precisions=i.precision(valid=True)[0][1]
    totalprecision.append(precisions)
c=statistics.mean(totalprecision)
c

0.9309923932750019


totalrecall = []
for i in models:
    recalls=i.recall(valid=True)[0][1]
    totalrecall.append(recalls)
d=statistics.mean(totalrecall)
d

Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.
Could not find exact threshold 0.0; using closest threshold found 0.0.

1.0


totalaccuracy = []
for i in models:
    accuracies=i.accuracy(valid=True)[0][1]
    totalaccuracy.append(accuracies)
e=statistics.mean(totalaccuracy)
e

0.9459681561376476


f1 = a
threshold = b
precision = c
sensitivity = d
accuracy = e
metrics = {'Threshold': threshold, 'F1': f1, 'Precision': precision, 'Sensitivity': sensitivity, 'Accuracy': accuracy}
gbm_performance = pd.DataFrame(metrics.values(), columns=['Value'], index=metrics.keys())
gbm_performance


gbm.varimp_plot()


start_time = pd.to_datetime('now') + pd.Timedelta('07:00:00')
models=[]
for train, test in zip(trains, tests):
    top_3vars = ['Card_Holder', 'Average_Transaction_Frequency', 'Merchant_ID']
    GLM = H2OAutoML(max_runtime_secs=10000, nfolds=30, include_algos=['GLM'])
    GLM.train(x=glm_vars, y='Fraud_Status', training_frame=train, validation_frame=test)
    models.append(GLM.leader)
timer(start_time, header='\nModeling')

---------------------------------------------------------------------------
ConnectionRefusedError                    Traceback (most recent call last)
/Applications/anaconda3/lib/python3.8/site-packages/urllib3/connection.py in _new_conn(self)
    168         try:
--> 169             conn = connection.create_connection(
    170                 (self._dns_host, self.port), self.timeout, **extra_kw

/Applications/anaconda3/lib/python3.8/site-packages/urllib3/util/connection.py in create_connection(address, timeout, source_address, socket_options)
     95     if err is not None:
---> 96         raise err
     97 

/Applications/anaconda3/lib/python3.8/site-packages/urllib3/util/connection.py in create_connection(address, timeout, source_address, socket_options)
     85                 sock.bind(source_address)
---> 86             sock.connect(sa)
     87             return sock

ConnectionRefusedError: [Errno 61] Connection refused

During handling of the above exception, another exception occurred:

NewConnectionError                        Traceback (most recent call last)
/Applications/anaconda3/lib/python3.8/site-packages/urllib3/connectionpool.py in urlopen(self, method, url, body, headers, retries, redirect, assert_same_host, timeout, pool_timeout, release_conn, chunked, body_pos, **response_kw)
    698             # Make the request on the httplib connection object.
--> 699             httplib_response = self._make_request(
    700                 conn,

/Applications/anaconda3/lib/python3.8/site-packages/urllib3/connectionpool.py in _make_request(self, conn, method, url, timeout, chunked, **httplib_request_kw)
    393             else:
--> 394                 conn.request(method, url, **httplib_request_kw)
    395 

/Applications/anaconda3/lib/python3.8/site-packages/urllib3/connection.py in request(self, method, url, body, headers)
    233             headers["User-Agent"] = _get_default_user_agent()
--> 234         super(HTTPConnection, self).request(method, url, body=body, headers=headers)
    235 

/Applications/anaconda3/lib/python3.8/http/client.py in request(self, method, url, body, headers, encode_chunked)
   1254         """Send a complete request to the server."""
-> 1255         self._send_request(method, url, body, headers, encode_chunked)
   1256 

/Applications/anaconda3/lib/python3.8/http/client.py in _send_request(self, method, url, body, headers, encode_chunked)
   1300             body = _encode(body, 'body')
-> 1301         self.endheaders(body, encode_chunked=encode_chunked)
   1302 

/Applications/anaconda3/lib/python3.8/http/client.py in endheaders(self, message_body, encode_chunked)
   1249             raise CannotSendHeader()
-> 1250         self._send_output(message_body, encode_chunked=encode_chunked)
   1251 

/Applications/anaconda3/lib/python3.8/http/client.py in _send_output(self, message_body, encode_chunked)
   1009         del self._buffer[:]
-> 1010         self.send(msg)
   1011 

/Applications/anaconda3/lib/python3.8/http/client.py in send(self, data)
    949             if self.auto_open:
--> 950                 self.connect()
    951             else:

/Applications/anaconda3/lib/python3.8/site-packages/urllib3/connection.py in connect(self)
    199     def connect(self):
--> 200         conn = self._new_conn()
    201         self._prepare_conn(conn)

/Applications/anaconda3/lib/python3.8/site-packages/urllib3/connection.py in _new_conn(self)
    180         except SocketError as e:
--> 181             raise NewConnectionError(
    182                 self, "Failed to establish a new connection: %s" % e

NewConnectionError: <urllib3.connection.HTTPConnection object at 0x7fc66fde3760>: Failed to establish a new connection: [Errno 61] Connection refused

During handling of the above exception, another exception occurred:

MaxRetryError                             Traceback (most recent call last)
/Applications/anaconda3/lib/python3.8/site-packages/requests/adapters.py in send(self, request, stream, timeout, verify, cert, proxies)
    438             if not chunked:
--> 439                 resp = conn.urlopen(
    440                     method=request.method,

/Applications/anaconda3/lib/python3.8/site-packages/urllib3/connectionpool.py in urlopen(self, method, url, body, headers, retries, redirect, assert_same_host, timeout, pool_timeout, release_conn, chunked, body_pos, **response_kw)
    754 
--> 755             retries = retries.increment(
    756                 method, url, error=e, _pool=self, _stacktrace=sys.exc_info()[2]

/Applications/anaconda3/lib/python3.8/site-packages/urllib3/util/retry.py in increment(self, method, url, response, error, _pool, _stacktrace)
    573         if new_retry.is_exhausted():
--> 574             raise MaxRetryError(_pool, url, error or ResponseError(cause))
    575 

MaxRetryError: HTTPConnectionPool(host='localhost', port=54321): Max retries exceeded with url: /3/Metadata/schemas/AutoMLV99 (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7fc66fde3760>: Failed to establish a new connection: [Errno 61] Connection refused'))

During handling of the above exception, another exception occurred:

ConnectionError                           Traceback (most recent call last)
/Applications/anaconda3/lib/python3.8/site-packages/h2o/backend/connection.py in request(self, endpoint, data, json, filename, save_to)
    471             verify = self._cacert if self._verify_ssl_cert and self._cacert else self._verify_ssl_cert
--> 472             resp = requests.request(method=method, url=url, data=data, json=json, files=files, params=params,
    473                                     headers=headers, timeout=self._timeout, stream=stream,

/Applications/anaconda3/lib/python3.8/site-packages/requests/api.py in request(method, url, **kwargs)
     60     with sessions.Session() as session:
---> 61         return session.request(method=method, url=url, **kwargs)
     62 

/Applications/anaconda3/lib/python3.8/site-packages/requests/sessions.py in request(self, method, url, params, data, headers, cookies, files, auth, timeout, allow_redirects, proxies, hooks, stream, verify, cert, json)
    541         send_kwargs.update(settings)
--> 542         resp = self.send(prep, **send_kwargs)
    543 

/Applications/anaconda3/lib/python3.8/site-packages/requests/sessions.py in send(self, request, **kwargs)
    654         # Send the request
--> 655         r = adapter.send(request, **kwargs)
    656 

/Applications/anaconda3/lib/python3.8/site-packages/requests/adapters.py in send(self, request, stream, timeout, verify, cert, proxies)
    515 
--> 516             raise ConnectionError(e, request=request)
    517 

ConnectionError: HTTPConnectionPool(host='localhost', port=54321): Max retries exceeded with url: /3/Metadata/schemas/AutoMLV99 (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7fc66fde3760>: Failed to establish a new connection: [Errno 61] Connection refused'))

During handling of the above exception, another exception occurred:

H2OConnectionError                        Traceback (most recent call last)
<ipython-input-197-61ebd3ca6477> in <module>
      3 for train, test in zip(trains, tests):
      4     top_3vars = ['Card_Holder', 'Average_Transaction_Frequency', 'Merchant_ID']
----> 5     GLM = H2OAutoML(max_runtime_secs=10000, nfolds=30, include_algos=['GLM'])
      6     GLM.train(x=glm_vars, y='Fraud_Status', training_frame=train, validation_frame=test)
      7     models.append(GLM.leader)

/Applications/anaconda3/lib/python3.8/site-packages/h2o/automl/autoh2o.py in __init__(self, nfolds, balance_classes, class_sampling_factors, max_after_balance_size, max_runtime_secs, max_runtime_secs_per_model, max_models, stopping_metric, stopping_tolerance, stopping_rounds, seed, project_name, exclude_algos, include_algos, exploitation_ratio, modeling_plan, preprocessing, monotone_constraints, keep_cross_validation_predictions, keep_cross_validation_models, keep_cross_validation_fold_assignment, sort_metric, export_checkpoints_dir, verbosity, **kwargs)
    153         # Check if H2O jar contains AutoML
    154         try:
--> 155             h2o.api("GET /3/Metadata/schemas/AutoMLV99")
    156         except h2o.exceptions.H2OResponseError as e:
    157             print(e)

/Applications/anaconda3/lib/python3.8/site-packages/h2o/h2o.py in api(endpoint, data, json, filename, save_to)
    105     # type checks are performed in H2OConnection class
    106     _check_connection()
--> 107     return h2oconn.request(endpoint, data=data, json=json, filename=filename, save_to=save_to)
    108 
    109 

/Applications/anaconda3/lib/python3.8/site-packages/h2o/backend/connection.py in request(self, endpoint, data, json, filename, save_to)
    484             else:
    485                 self._log_end_exception(e)
--> 486                 raise H2OConnectionError("Unexpected HTTP error: %s" % e)
    487         except requests.exceptions.Timeout as e:
    488             self._log_end_exception(e)

H2OConnectionError: Unexpected HTTP error: HTTPConnectionPool(host='localhost', port=54321): Max retries exceeded with url: /3/Metadata/schemas/AutoMLV99 (Caused by NewConnectionError('<urllib3.connection.HTTPConnection object at 0x7fc66fde3760>: Failed to establish a new connection: [Errno 61] Connection refused'))


glm_3vars

Model Details
=============
H2OGeneralizedLinearEstimator :  Generalized Linear Modeling
Model Key:  GLM_1_AutoML_20210701_182845


GLM Model: summary


ModelMetricsBinomialGLM: glm
** Reported on train data. **

MSE: 0.04909037783131847
RMSE: 0.2215634848780784
LogLoss: 0.19104180709294746
Null degrees of freedom: 7785
Residual degrees of freedom: 7782
Null deviance: 3538.959233793834
Residual deviance: 2974.9030200513785
AIC: 2982.9030200513785
AUC: 0.7718163035531168
AUCPR: 0.2793035805825629
Gini: 0.5436326071062336

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.25003107330309315:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  6.01 %


ModelMetricsBinomialGLM: glm
** Reported on validation data. **

MSE: 0.04715765970836517
RMSE: 0.2171581444670339
LogLoss: 0.18324846751149748
Null degrees of freedom: 1946
Residual degrees of freedom: 1943
Null deviance: 884.801798631153
Residual deviance: 713.5695324897711
AIC: 721.5695324897711
AUC: 0.7501144271636074
AUCPR: 0.30288900816332487
Gini: 0.5002288543272149

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.08655307529027566:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  6.14 %


ModelMetricsBinomialGLM: glm
** Reported on cross-validation data. **

MSE: 0.04918396499714938
RMSE: 0.2217745814947001
LogLoss: 0.19150429178700995
Null degrees of freedom: 7785
Residual degrees of freedom: 7782
Null deviance: 3540.9687371308805
Residual deviance: 2982.104831707319
AIC: 2990.104831707319
AUC: 0.766707865864056
AUCPR: 0.2700517998881614
Gini: 0.5334157317281121

Confusion Matrix (Act/Pred) for max f1 @ threshold = 0.24807023134781647:

Maximum Metrics: Maximum metrics at their respective thresholds

Gains/Lift Table: Avg response rate:  6.01 %, avg score:  6.01 %


totalF1 = []
for i in models:
    F1s=i.F1(valid=True)[0][1]
    totalF1.append(F1s)
a=statistics.mean(totalF1)
a


totalthreshold = []
for i in models:
    thresholds=i.F1(valid=True)[0][0]
    totalthreshold.append(thresholds)
b=statistics.mean(totalthreshold)
b


totalprecision = []
for i in models:
    precisions=i.precision(valid=True)[0][1]
    totalprecision.append(precisions)
c=statistics.mean(totalprecision)
c


totalrecall = []
for i in models:
    recalls=i.recall(valid=True)[0][1]
    totalrecall.append(recalls)
d=statistics.mean(totalrecall)
d


totalaccuracy = []
for i in models:
    accuracies=i.accuracy(valid=True)[0][1]
    totalaccuracy.append(accuracies)
e=statistics.mean(totalaccuracy)
e


f1 = a
threshold = b
precision = c
sensitivity = d
accuracy = e
metrics = {'Threshold': threshold, 'F1': f1, 'Precision': precision, 'Sensitivity': sensitivity, 'Accuracy': accuracy}
glm_3vars_performance = pd.DataFrame(metrics.values(), columns=['Value'], index=metrics.keys())
glm_3vars_performance


performances = [glm_performance, gbm_performance, glm_3vars_performance]
model_performances = pd.concat(performances, axis=1)
model_performances


filepath = r'~/Desktop/Desktop/Prasmul/Semester 6/Applied Data Science for Business/Project Group Dataset/'
filename = r'cctx_test.xlsx'

datatest = pd.read_excel(filepath + filename)
datatest.head()


datatest.info(memory_usage='deep')

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2625 entries, 0 to 2624
Data columns (total 20 columns):
 #   Column                         Non-Null Count  Dtype         
---  ------                         --------------  -----         
 0   Transaction_ID                 2625 non-null   object        
 1   Transaction_Flag               2625 non-null   bool          
 2   Transaction_Date               2625 non-null   datetime64[ns]
 3   Transaction_Type               2625 non-null   object        
 4   Transaction_Amount             2625 non-null   int64         
 5   Bank_ID                        2625 non-null   int64         
 6   CC_ID                          2625 non-null   object        
 7   Card_Type                      2625 non-null   object        
 8   Card_Holder                    2625 non-null   int64         
 9   Channel_ID                     2625 non-null   int64         
 10  Merchant_ID                    2625 non-null   object        
 11  Country_ID                     2625 non-null   object        
 12  City_ID                        2625 non-null   object        
 13  EDC_Type                       2625 non-null   object        
 14  EDC_Location                   2625 non-null   object        
 15  EDC_Owner                      2625 non-null   object        
 16  Average_Transaction_Amount     2620 non-null   float64       
 17  Maximum_Transaction_Amount     2620 non-null   float64       
 18  Minimum_Transaction_Amount     2620 non-null   float64       
 19  Average_Transaction_Frequency  2620 non-null   float64       
dtypes: bool(1), datetime64[ns](1), float64(4), int64(4), object(10)
memory usage: 1.8 MB


def null_counts(df, style=True):
    nulls = df.isna().sum().rename_axis('Columns').reset_index(name='Count')
    nulls['Percentage'] = nulls['Count'] / len(df)
    nulls = nulls.loc[nulls['Count'] > 0]
    nulls.sort_values(by='Count', ascending=False, inplace=True)
    nulls.reset_index(drop=True, inplace=True)
    
    if style:
        nulls = nulls.style.format({'Count': '{:,}', 'Percentage': '{:.2%}'}).hide_index()
        
    return nulls


null_counts(datatest)


datatest.dropna()


datatest.describe()


datatest.describe(include=['O'])


datatest = datatest.drop(['Transaction_Flag','Country_ID','Bank_ID','CC_ID','Transaction_Date','EDC_Location'], axis=1)
datatest.head()


datatest['Card_Type'].replace({'CC11': 0, 'CC08': 1, 'CC09':2, 'CC00':3, 'CC10': 4}, inplace=True)
unused = datatest['Card_Type'].loc[(datatest['Card_Type']!= 0)&(datatest['Card_Type']!=1)& (datatest['Card_Type']!=2) & (datatest['Card_Type']!=3) &(datatest['Card_Type']!=4)]
datatest.replace(unused.values, 5, inplace=True)
datatest


datatest['Transaction_Type'].replace({'T01': 0, 'T15': 1, 'T02':2, 'T08':3}, inplace=True)
unused = datatest['Transaction_Type'].loc[(datatest['Transaction_Type']!= 0)&(datatest['Transaction_Type']!=1)& (datatest['Transaction_Type']!=2) & (datatest['Transaction_Type']!=3)]
datatest.replace(unused.values, 4, inplace=True)
datatest


datatest['Card_Holder'].replace({1: 0, 2: 1}, inplace=True)


datatest['EDC_Type'].replace({'EDC0002':0}, inplace=True)
unused = datatest['EDC_Type'].loc[datatest['EDC_Type']!= 0]
datatest.replace(unused.values, 1, inplace=True)
datatest


datatest['Channel_ID'].replace({1:1}, inplace=True)
unused = datatest['Channel_ID'].loc[datatest['Channel_ID']!= 1]
datatest.replace(unused.values, 0, inplace=True)
datatest


datatest['Merchant_ID'].replace({'M0001':1}, inplace=True)
unused = datatest['Merchant_ID'].loc[datatest['Merchant_ID']!= 1]
datatest.replace(unused.values, 0, inplace=True)
datatest


datatest['City_ID'].replace({'CTY06-023':0}, inplace=True)
unused = datatest['City_ID'].loc[datatest['City_ID']!= 0]
datatest.replace(unused.values, 1, inplace=True)
datatest


datatest['EDC_Owner'].replace({'OEDC0377':1}, inplace=True)
unused = datatest['EDC_Owner'].loc[datatest['EDC_Owner']!= 1]
datatest.replace(unused.values, 0, inplace=True)
datatest


datatest.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 2625 entries, 0 to 2624
Data columns (total 14 columns):
 #   Column                         Non-Null Count  Dtype  
---  ------                         --------------  -----  
 0   Transaction_ID                 2625 non-null   object 
 1   Transaction_Type               2625 non-null   int64  
 2   Transaction_Amount             2625 non-null   int64  
 3   Card_Type                      2625 non-null   int64  
 4   Card_Holder                    2625 non-null   int64  
 5   Channel_ID                     2625 non-null   int64  
 6   Merchant_ID                    2625 non-null   int64  
 7   City_ID                        2625 non-null   int64  
 8   EDC_Type                       2625 non-null   int64  
 9   EDC_Owner                      2625 non-null   int64  
 10  Average_Transaction_Amount     2620 non-null   float64
 11  Maximum_Transaction_Amount     2620 non-null   float64
 12  Minimum_Transaction_Amount     2620 non-null   float64
 13  Average_Transaction_Frequency  2620 non-null   float64
dtypes: float64(4), int64(9), object(1)
memory usage: 287.2+ KB


datatest.reset_index(drop=True, inplace=True)
datatest


datafinal = h2o.H2OFrame(datatest)

Parse progress: |█████████████████████████████████████████████████████████| 100%


predict = glm_3vars.predict(datafinal)

glm prediction progress: |████████████████████████████████████████████████| 100%


data_glm = h2o.as_list(predict, use_pandas=True)


data_glm


datatest['Fraud_Status']=data_glm['predict']


final_prediction = datatest[['Transaction_ID', 'Fraud_Status']]


final_prediction


final_prediction.to_csv('final_prediction.csv')

	Transaction_Amount	Bank_ID	Card_Holder	Channel_ID	Average_Transaction_Amount	Maximum_Transaction_Amount	Minimum_Transaction_Amount	Average_Transaction_Frequency	Fraud_Status
count	1.050000e+04	10500.0	10500.000000	10500.000000	1.048400e+04	1.048400e+04	1.048400e+04	10484.000000	10500.000000
mean	1.320409e+06	1.0	1.932190	1.405333	1.366425e+06	1.229132e+07	7.735985e+04	2.435969	0.069333
std	2.819677e+06	0.0	0.251431	1.013994	1.439952e+06	1.640702e+07	7.496893e+05	1.390293	0.254032
min	1.000000e+00	1.0	1.000000	1.000000	5.000000e+04	3.800000e+04	1.000000e+00	1.000000	0.000000
25%	2.000000e+05	1.0	2.000000	1.000000	5.708772e+05	2.500000e+06	2.500000e+04	1.680000	0.000000
50%	5.800000e+05	1.0	2.000000	1.000000	1.026553e+06	6.000000e+06	3.750000e+04	2.100000	0.000000
75%	1.250000e+06	1.0	2.000000	1.000000	1.691944e+06	1.500000e+07	6.550000e+04	2.780000	0.000000
max	6.000000e+07	1.0	2.000000	5.000000	2.466667e+07	1.000000e+08	7.500000e+07	19.780000	1.000000

Value	Count	Percentage
CC11	3,841	36.58%
CC08	2,113	20.12%
CC09	1,381	13.15%
CC00	1,260	12.00%
CC10	1,093	10.41%
CC04	237	2.26%
CC02	157	1.50%
CC01	155	1.48%
CC05	145	1.38%
CC12	54	0.51%
CC13	36	0.34%
CC03	20	0.19%
CC06	7	0.07%
CC07	1	0.01%

Value	Count	Percentage
T01	2,872	27.35%
T15	2,449	23.32%
T02	1,324	12.61%
T08	1,074	10.23%
T13	769	7.32%
T06	677	6.45%
T14	509	4.85%
T04	264	2.51%
T03	207	1.97%
T10	112	1.07%
T12	77	0.73%
T11	50	0.48%
T20	29	0.28%
T17	27	0.26%
T05	22	0.21%
T16	12	0.11%
T09	8	0.08%
T07	7	0.07%
T19	7	0.07%
T18	4	0.04%

Value	Count	Percentage
2	9,788	93.22%
1	712	6.78%

	Transaction_Date	Transaction_Type	Transaction_Amount	Card_Type	Card_Holder	Channel_ID	Merchant_ID	City_ID	EDC_Type	EDC_Owner	Average_Transaction_Amount	Maximum_Transaction_Amount	Minimum_Transaction_Amount	Average_Transaction_Frequency	Fraud_Status
0	2018-01-01 01:48:50.951	3	50000.0	0	1	1	M0001	CTY06-133	1	OEDC0377	661855.03	11000000.0	24212.0	2.91	0
1	2018-01-01 09:08:52.666	0	1000000.0	1	1	1	M0001	CTY06-129	1	OEDC0377	2171907.10	28555000.0	100000.0	3.12	1
2	2018-01-01 09:08:52.666	1	1000000.0	1	1	1	M0001	CTY06-073	1	OEDC0377	1714437.98	8500000.0	50.0	3.58	0
3	2018-01-01 09:45:55.969	1	1000000.0	1	1	1	M0001	CTY06-129	1	OEDC0377	10163708.23	100000000.0	63000.0	2.57	0
4	2018-01-01 23:41:59.228	4	2500000.0	5	1	1	M0001	CTY06-133	1	OEDC0377	690066.65	3523000.0	26500.0	4.23	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
10495	2018-12-31 15:47:34.782	4	5000000.0	1	1	1	M0001	CTY06-023	1	OEDC0377	1667486.67	20000000.0	50000.0	4.15	0
10496	2018-12-31 15:47:34.782	4	4800000.0	0	1	1	M0001	CTY06-181	1	OEDC0377	2530285.33	5000000.0	1000000.0	1.64	0
10497	2018-12-31 15:47:34.782	4	2500000.0	3	1	1	M0001	CTY06-023	1	OEDC0377	1225000.00	10000000.0	100000.0	1.97	0
10498	2018-12-31 15:47:34.782	4	1100000.0	0	1	1	M0001	CTY06-186	1	OEDC0377	820036.79	4000000.0	50000.0	3.16	0
10499	2018-12-31 22:11:05.961	2	102500.0	0	1	2	M0597	CTY06-171	1	OEDC0633	571282.15	5000000.0	25000.0	2.36	0

Import Packages and Data¶

Data Cleansing¶

Checking Target Balance.¶

Data Preparation¶

Explanatory Analysis¶

Building Model¶

Base Model (GLM)¶

Random Forest¶

GLM 3 Vars¶

Test Data¶

Data Preparation¶

	Transaction_ID	Transaction_Flag	Transaction_Date	Transaction_Type	Transaction_Amount	Bank_ID	CC_ID	Card_Type	Card_Holder	Channel_ID	Merchant_ID	Country_ID	City_ID	EDC_Type	EDC_Location	EDC_Owner	Average_Transaction_Amount	Maximum_Transaction_Amount	Minimum_Transaction_Amount	Average_Transaction_Frequency	Fraud_Status
0	TX00002	False	2018-01-01 01:48:50.951	T08	50000.0	1	CCID5563	CC11	2	1	M0001	CTY06	CTY06-133	EDC0885	LEDC3703	OEDC0377	661855.03	11000000.0	24212.0	2.91	0
1	TX00004	False	2018-01-01 09:08:52.666	T01	1000000.0	1	CCID4598	CC08	2	1	M0001	CTY06	CTY06-129	EDC0565	LEDC3205	OEDC0377	2171907.10	28555000.0	100000.0	3.12	1
2	TX00005	False	2018-01-01 09:08:52.666	T15	1000000.0	1	CCID2839	CC08	2	1	M0001	CTY06	CTY06-073	EDC4639	LEDC3081	OEDC0377	1714437.98	8500000.0	50.0	3.58	0
3	TX00006	False	2018-01-01 09:45:55.969	T15	1000000.0	1	CCID2968	CC08	2	1	M0001	CTY06	CTY06-129	EDC3918	LEDC1993	OEDC0377	10163708.23	100000000.0	63000.0	2.57	0
4	TX00007	False	2018-01-01 23:41:59.228	T10	2500000.0	1	CCID0176	CC12	2	1	M0001	CTY06	CTY06-133	EDC2863	LEDC2062	OEDC0377	690066.65	3523000.0	26500.0	4.23	0

Columns	Count	Percentage
Average_Transaction_Amount	16	0.15%
Maximum_Transaction_Amount	16	0.15%
Minimum_Transaction_Amount	16	0.15%
Average_Transaction_Frequency	16	0.15%

	Transaction_ID	Transaction_Type	CC_ID	Card_Type	Merchant_ID	Country_ID	City_ID	EDC_Type	EDC_Location	EDC_Owner
count	10500	10500	10500	10500	10500	10500	10500	10500	10500	10500
unique	10500	20	7573	14	975	11	227	4789	5238	1461
top	TX02634	T01	CCID2268	CC11	M0001	CTY06	CTY06-023	EDC0002	LEDC0400	OEDC0377
freq	1	2872	5	3841	9037	10464	4255	707	16	8325

Value	Count	Percentage
M0001	9,037	86.07%
M0243	6	0.06%
M0140	6	0.06%
M0063	6	0.06%
M0100	6	0.06%
M0343	5	0.05%
M0405	5	0.05%
M0088	5	0.05%
M0187	5	0.05%
M0307	4	0.04%
M0592	4	0.04%
M0418	4	0.04%
M0200	4	0.04%
M0514	4	0.04%
M0633	4	0.04%
M0013	4	0.04%
M0323	4	0.04%
M0104	4	0.04%
M0156	4	0.04%
M0510	4	0.04%
M0081	4	0.04%
M0419	4	0.04%
M0644	4	0.04%
M0325	4	0.04%
M0350	4	0.04%
M0434	4	0.04%
M0208	4	0.04%
M0155	4	0.04%
M0202	4	0.04%
M0649	4	0.04%
M0017	4	0.04%
M0524	3	0.03%
M0447	3	0.03%
M0111	3	0.03%
M0117	3	0.03%
M0391	3	0.03%
M0093	3	0.03%
M0486	3	0.03%
M0274	3	0.03%
M0363	3	0.03%
M0084	3	0.03%
M0332	3	0.03%
M0437	3	0.03%
M0231	3	0.03%
M0511	3	0.03%
M0004	3	0.03%
M0463	3	0.03%
M0220	3	0.03%
M0618	3	0.03%
M0087	3	0.03%
M0267	3	0.03%
M0019	3	0.03%
M0234	3	0.03%
M0354	3	0.03%
M0523	3	0.03%
M0472	3	0.03%
M0153	3	0.03%
M0708	3	0.03%
M0059	3	0.03%
M0287	3	0.03%
M0014	3	0.03%
M0184	3	0.03%
M0545	3	0.03%
M0394	3	0.03%
M0058	3	0.03%
M0590	3	0.03%
M0549	3	0.03%
M0238	3	0.03%
M0565	3	0.03%
M0397	3	0.03%
M0393	3	0.03%
M0724	3	0.03%
M0168	3	0.03%
M0719	3	0.03%
M0544	3	0.03%
M0326	3	0.03%
M0141	3	0.03%
M0169	3	0.03%
M0538	3	0.03%
M0497	3	0.03%
M0008	3	0.03%
M0114	3	0.03%
M0294	3	0.03%
M0221	3	0.03%
M0467	3	0.03%
M0310	3	0.03%
M0262	3	0.03%
M0075	3	0.03%
M0588	3	0.03%
M0159	3	0.03%
M0424	3	0.03%
M0313	3	0.03%
M0413	3	0.03%
M0139	3	0.03%
M0776	3	0.03%
M0261	3	0.03%
M0764	3	0.03%
M0475	3	0.03%
M0714	3	0.03%
M0229	3	0.03%
M0687	3	0.03%
M0493	3	0.03%
M0715	3	0.03%
M0259	3	0.03%
M0197	3	0.03%
M0595	3	0.03%
M0745	3	0.03%
M0038	3	0.03%
M0795	3	0.03%
M0006	3	0.03%
M0382	3	0.03%
M0386	3	0.03%
M0839	3	0.03%
M0358	3	0.03%
M0657	3	0.03%
M0282	3	0.03%
M1025	2	0.02%
M0828	2	0.02%
M0183	2	0.02%
M0039	2	0.02%
M0620	2	0.02%
M0627	2	0.02%
M0449	2	0.02%
M0384	2	0.02%
M0471	2	0.02%
M0850	2	0.02%
M0872	2	0.02%
M0789	2	0.02%
M0579	2	0.02%
M0025	2	0.02%
M0022	2	0.02%
M1007	2	0.02%
M0666	2	0.02%
M0285	2	0.02%
M0409	2	0.02%
M0534	2	0.02%
M0451	2	0.02%
M0669	2	0.02%
M0533	2	0.02%
M0406	2	0.02%
M0757	2	0.02%
M0580	2	0.02%
M0360	2	0.02%
M0042	2	0.02%
M0346	2	0.02%
M0170	2	0.02%
M0124	2	0.02%
M0547	2	0.02%
M0468	2	0.02%
M0148	2	0.02%
M0302	2	0.02%
M0816	2	0.02%
M0143	2	0.02%
M0045	2	0.02%
M0304	2	0.02%
M0503	2	0.02%
M0256	2	0.02%
M0309	2	0.02%
M0665	2	0.02%
M0494	2	0.02%
M0348	2	0.02%
M0893	2	0.02%
M0837	2	0.02%
M0341	2	0.02%
M0536	2	0.02%
M0732	2	0.02%
M0048	2	0.02%
M0272	2	0.02%
M0833	2	0.02%
M0178	2	0.02%
M0366	2	0.02%
M0530	2	0.02%
M0237	2	0.02%
M0336	2	0.02%
M0079	2	0.02%
M0012	2	0.02%
M0277	2	0.02%
M0853	2	0.02%
M0268	2	0.02%
M0137	2	0.02%
M0314	2	0.02%
M0328	2	0.02%
M0445	2	0.02%
M0191	2	0.02%
M0072	2	0.02%
M0320	2	0.02%
M0851	2	0.02%
M0707	2	0.02%
M0790	2	0.02%
M0086	2	0.02%
M0185	2	0.02%
M0319	2	0.02%
M0249	2	0.02%
M0367	2	0.02%
M1084	2	0.02%
M0980	2	0.02%
M0308	2	0.02%
M0298	2	0.02%
M0611	2	0.02%
M0832	2	0.02%
M0353	2	0.02%
M0501	2	0.02%
M0110	2	0.02%
M0689	2	0.02%
M0891	2	0.02%
M0009	2	0.02%
M0462	2	0.02%
M0404	2	0.02%
M0216	2	0.02%
M0922	2	0.02%
M0067	2	0.02%
M0460	2	0.02%
M0821	2	0.02%
M0123	2	0.02%
M0563	2	0.02%
M0160	2	0.02%
M0091	2	0.02%
M0635	2	0.02%
M0513	2	0.02%
M1012	2	0.02%
M0619	2	0.02%
M0603	2	0.02%
M0988	2	0.02%
M1060	2	0.02%
M0147	2	0.02%
M0292	2	0.02%
M0134	2	0.02%
M0018	2	0.02%
M0352	2	0.02%
M0260	2	0.02%
M0275	2	0.02%
M0071	2	0.02%
M0650	2	0.02%
M0317	2	0.02%
M0648	2	0.02%
M0257	2	0.02%
M0543	2	0.02%
M0196	2	0.02%
M0312	2	0.02%
M0696	2	0.02%
M0504	2	0.02%
M0740	2	0.02%
M1049	2	0.02%
M0246	2	0.02%
M0454	2	0.02%
M0604	2	0.02%
M0151	2	0.02%
M0021	2	0.02%
M0333	2	0.02%
M0975	2	0.02%
M0263	2	0.02%
M0456	2	0.02%
M0654	2	0.02%
M0609	2	0.02%
M0699	2	0.02%
M0372	2	0.02%
M0817	2	0.02%
M0023	2	0.02%
M0024	2	0.02%
M0094	2	0.02%
M0852	2	0.02%
M0450	2	0.02%
M0967	2	0.02%
M0729	2	0.02%
M0421	2	0.02%
M0917	2	0.02%
M0435	2	0.02%
M0062	2	0.02%
M0171	2	0.02%
M0632	2	0.02%
M0509	2	0.02%
M0057	2	0.02%
M0797	2	0.02%
M0517	2	0.02%
M0158	2	0.02%
M0389	2	0.02%
M0327	2	0.02%
M0402	2	0.02%
M0379	2	0.02%
M0481	2	0.02%
M0181	2	0.02%
M0743	2	0.02%
M0783	2	0.02%
M0571	2	0.02%
M0362	2	0.02%
M0663	2	0.02%
M0875	2	0.02%
M1022	2	0.02%
M0315	2	0.02%
M0780	2	0.02%
M0361	2	0.02%
M0845	2	0.02%
M0281	2	0.02%
M0559	2	0.02%
M0505	2	0.02%
M0702	2	0.02%
M0218	2	0.02%
M0541	2	0.02%
M0280	2	0.02%
M0428	2	0.02%
M0020	2	0.02%
M0398	2	0.02%
M0005	2	0.02%
M0484	2	0.02%
M0115	2	0.02%
M0677	2	0.02%
M0801	2	0.02%
M0763	2	0.02%
M0121	2	0.02%
M0690	2	0.02%
M0756	2	0.02%
M0892	2	0.02%
M0112	2	0.02%
M0193	2	0.02%
M0125	2	0.02%
M0368	2	0.02%
M0299	2	0.02%
M0936	2	0.02%
M0754	2	0.02%
M0383	2	0.02%
M0050	2	0.02%
M0829	2	0.02%
M0070	2	0.02%
M0250	2	0.02%
M0189	2	0.02%
M0138	2	0.02%
M0876	2	0.02%
M0805	2	0.02%
M0396	2	0.02%
M0378	2	0.02%
M0417	2	0.02%
M0136	2	0.02%
M0515	2	0.02%
M1114	1	0.01%
M1039	1	0.01%
M0999	1	0.01%
M0520	1	0.01%
M0064	1	0.01%
M1109	1	0.01%
M0190	1	0.01%
M0264	1	0.01%
M0629	1	0.01%
M0295	1	0.01%
M0914	1	0.01%
M0920	1	0.01%
M0926	1	0.01%
M0203	1	0.01%
M0930	1	0.01%
M0660	1	0.01%
M0791	1	0.01%
M0992	1	0.01%
M0150	1	0.01%
M0375	1	0.01%
M0906	1	0.01%
M0862	1	0.01%
M0923	1	0.01%
M0997	1	0.01%
M0452	1	0.01%
M0728	1	0.01%
M1048	1	0.01%
M0955	1	0.01%
M0116	1	0.01%
M0551	1	0.01%
M0761	1	0.01%
M0961	1	0.01%
M0010	1	0.01%
M0126	1	0.01%
M0427	1	0.01%
M0446	1	0.01%
M0003	1	0.01%
M0399	1	0.01%
M0172	1	0.01%
M0877	1	0.01%
M0400	1	0.01%
M1035	1	0.01%
M0583	1	0.01%
M0624	1	0.01%
M0374	1	0.01%
M0971	1	0.01%
M0830	1	0.01%
M0179	1	0.01%
M0502	1	0.01%
M0132	1	0.01%
M0940	1	0.01%
M0972	1	0.01%
M0258	1	0.01%
M0807	1	0.01%
M0737	1	0.01%
M0739	1	0.01%
M0297	1	0.01%
M0469	1	0.01%
M0466	1	0.01%
M0786	1	0.01%
M0356	1	0.01%
M0929	1	0.01%
M0029	1	0.01%
M0192	1	0.01%
M0482	1	0.01%
M0919	1	0.01%
M0918	1	0.01%
M0276	1	0.01%
M0802	1	0.01%
M0686	1	0.01%
M1013	1	0.01%
M1004	1	0.01%
M0823	1	0.01%
M0762	1	0.01%
M0403	1	0.01%
M0230	1	0.01%
M1085	1	0.01%
M0144	1	0.01%
M0080	1	0.01%
M0431	1	0.01%
M0458	1	0.01%
M0827	1	0.01%
M0641	1	0.01%
M0645	1	0.01%
M0938	1	0.01%
M0182	1	0.01%
M0443	1	0.01%
M0478	1	0.01%
M0300	1	0.01%
M1119	1	0.01%
M0364	1	0.01%
M0895	1	0.01%
M0002	1	0.01%
M0810	1	0.01%
M0573	1	0.01%
M0718	1	0.01%
M0046	1	0.01%
M1094	1	0.01%
M0539	1	0.01%
M1040	1	0.01%
M0291	1	0.01%
M0681	1	0.01%
M0751	1	0.01%
M0808	1	0.01%
M0826	1	0.01%
M0857	1	0.01%
M0388	1	0.01%
M0792	1	0.01%
M0990	1	0.01%
M0924	1	0.01%
M0752	1	0.01%
M0978	1	0.01%
M0744	1	0.01%
M1073	1	0.01%
M0723	1	0.01%
M0149	1	0.01%
M0215	1	0.01%
M0154	1	0.01%
M1058	1	0.01%
M0480	1	0.01%
M0305	1	0.01%
M0847	1	0.01%
M0052	1	0.01%
M0896	1	0.01%
M0376	1	0.01%
M0954	1	0.01%
M0240	1	0.01%
M0741	1	0.01%
M0145	1	0.01%
M0390	1	0.01%
M1082	1	0.01%
M0638	1	0.01%
M0835	1	0.01%
M0491	1	0.01%
M0442	1	0.01%
M0271	1	0.01%
M0204	1	0.01%
M0630	1	0.01%
M0054	1	0.01%
M0564	1	0.01%
M1020	1	0.01%
M0765	1	0.01%
M0040	1	0.01%
M0772	1	0.01%
M0444	1	0.01%
M0567	1	0.01%
M0357	1	0.01%
M0490	1	0.01%
M1052	1	0.01%
M0041	1	0.01%
M0107	1	0.01%
M0371	1	0.01%
M0927	1	0.01%
M0788	1	0.01%
M0911	1	0.01%
M0198	1	0.01%
M0982	1	0.01%
M0908	1	0.01%
M1032	1	0.01%
M0099	1	0.01%
M0974	1	0.01%
M0127	1	0.01%
M0670	1	0.01%
M0948	1	0.01%
M0613	1	0.01%
M0537	1	0.01%
M0207	1	0.01%
M0912	1	0.01%
M1077	1	0.01%
M0993	1	0.01%
M0774	1	0.01%
M0146	1	0.01%
M0252	1	0.01%
M0606	1	0.01%
M0683	1	0.01%
M0652	1	0.01%
M0984	1	0.01%
M0864	1	0.01%
M0453	1	0.01%
M0129	1	0.01%
M0301	1	0.01%
M0639	1	0.01%
M1099	1	0.01%
M0848	1	0.01%
M0889	1	0.01%
M0037	1	0.01%
M0903	1	0.01%
M0199	1	0.01%
M0697	1	0.01%
M0904	1	0.01%
M0102	1	0.01%
M0284	1	0.01%
M0722	1	0.01%
M0642	1	0.01%
M0420	1	0.01%
M1087	1	0.01%
M0142	1	0.01%
M1104	1	0.01%
M0508	1	0.01%
M0939	1	0.01%
M0615	1	0.01%
M0700	1	0.01%
M1029	1	0.01%
M0130	1	0.01%
M1033	1	0.01%
M0858	1	0.01%
M0066	1	0.01%
M0996	1	0.01%
M0255	1	0.01%
M1090	1	0.01%
M0174	1	0.01%
M1000	1	0.01%
M0748	1	0.01%
M0043	1	0.01%
M0679	1	0.01%
M0989	1	0.01%
M0245	1	0.01%
M1095	1	0.01%
M0076	1	0.01%
M0560	1	0.01%
M0986	1	0.01%
M0881	1	0.01%
M1070	1	0.01%
M1046	1	0.01%
M0173	1	0.01%
M0778	1	0.01%
M0658	1	0.01%
M0152	1	0.01%
M0878	1	0.01%
M0843	1	0.01%
M0931	1	0.01%
M0928	1	0.01%
M0550	1	0.01%
M0128	1	0.01%
M0899	1	0.01%
M0555	1	0.01%
M0860	1	0.01%
M0278	1	0.01%
M0531	1	0.01%
M0213	1	0.01%
M0733	1	0.01%
M1072	1	0.01%
M0118	1	0.01%
M0959	1	0.01%
M0440	1	0.01%
M0587	1	0.01%
M0026	1	0.01%
M0883	1	0.01%
M0568	1	0.01%
M0167	1	0.01%
M1103	1	0.01%
M0098	1	0.01%
M1089	1	0.01%
M0570	1	0.01%
M0594	1	0.01%
M0626	1	0.01%
M0056	1	0.01%
M0758	1	0.01%
M0849	1	0.01%
M0869	1	0.01%
M1001	1	0.01%
M1102	1	0.01%
M0597	1	0.01%
M0991	1	0.01%
M0101	1	0.01%
M0736	1	0.01%
M1056	1	0.01%
M0651	1	0.01%
M1042	1	0.01%
M0809	1	0.01%
M0672	1	0.01%
M0621	1	0.01%
M0656	1	0.01%
M0279	1	0.01%
M0032	1	0.01%
M0664	1	0.01%
M1024	1	0.01%
M0387	1	0.01%
M0734	1	0.01%
M1106	1	0.01%
M0122	1	0.01%
M0188	1	0.01%
M0882	1	0.01%
M0866	1	0.01%
M0814	1	0.01%
M0953	1	0.01%
M0210	1	0.01%
M0429	1	0.01%
M0251	1	0.01%
M0035	1	0.01%
M0316	1	0.01%
M0177	1	0.01%
M0987	1	0.01%
M0863	1	0.01%
M0945	1	0.01%
M0769	1	0.01%
M0610	1	0.01%
M0244	1	0.01%
M0907	1	0.01%
M0488	1	0.01%
M0749	1	0.01%
M0865	1	0.01%
M0365	1	0.01%
M0951	1	0.01%
M0634	1	0.01%
M0868	1	0.01%
M0254	1	0.01%
M0526	1	0.01%
M0921	1	0.01%
M0653	1	0.01%
M0235	1	0.01%
M0455	1	0.01%
M0438	1	0.01%
M0355	1	0.01%
M0915	1	0.01%
M0693	1	0.01%
M0785	1	0.01%
M0288	1	0.01%
M0870	1	0.01%
M1076	1	0.01%
M0591	1	0.01%
M0507	1	0.01%
M0496	1	0.01%
M1100	1	0.01%
M0339	1	0.01%
M0532	1	0.01%
M0109	1	0.01%
M0119	1	0.01%
M0036	1	0.01%
M0819	1	0.01%
M1080	1	0.01%
M0824	1	0.01%
M0694	1	0.01%
M0750	1	0.01%
M0960	1	0.01%
M0871	1	0.01%
M1068	1	0.01%
M0554	1	0.01%
M0958	1	0.01%
M0521	1	0.01%
M0499	1	0.01%
M0947	1	0.01%
M0558	1	0.01%
M0226	1	0.01%
M0219	1	0.01%
M0436	1	0.01%
M0582	1	0.01%
M1037	1	0.01%
M0422	1	0.01%
M0248	1	0.01%
M0028	1	0.01%
M1113	1	0.01%
M0027	1	0.01%
M0676	1	0.01%
M0448	1	0.01%
M0596	1	0.01%
M0525	1	0.01%
M0831	1	0.01%
M0569	1	0.01%
M0969	1	0.01%
M0804	1	0.01%
M0901	1	0.01%
M1066	1	0.01%
M0519	1	0.01%
M0483	1	0.01%
M0753	1	0.01%
M0175	1	0.01%
M1081	1	0.01%
M1098	1	0.01%
M0956	1	0.01%
M0576	1	0.01%
M0738	1	0.01%
M0735	1	0.01%
M0796	1	0.01%
M0622	1	0.01%
M0760	1	0.01%
M1047	1	0.01%
M0330	1	0.01%
M1050	1	0.01%
M0759	1	0.01%
M0131	1	0.01%
M0572	1	0.01%
M0646	1	0.01%
M0381	1	0.01%
M0706	1	0.01%
M0540	1	0.01%
M0779	1	0.01%
M0495	1	0.01%
M0767	1	0.01%
M1015	1	0.01%
M0423	1	0.01%
M0522	1	0.01%
M0007	1	0.01%
M0680	1	0.01%
M0888	1	0.01%
M0324	1	0.01%
M0073	1	0.01%
M1036	1	0.01%
M0973	1	0.01%
M1059	1	0.01%
M0338	1	0.01%
M1023	1	0.01%
M1111	1	0.01%
M1043	1	0.01%
M0701	1	0.01%
M0518	1	0.01%
M0584	1	0.01%
M0205	1	0.01%
M0887	1	0.01%
M0946	1	0.01%
M0623	1	0.01%
M1044	1	0.01%
M0082	1	0.01%
M0682	1	0.01%
M1064	1	0.01%
M0068	1	0.01%
M0269	1	0.01%
M0306	1	0.01%
M0212	1	0.01%
M0180	1	0.01%
M0461	1	0.01%
M0726	1	0.01%
M1122	1	0.01%
M0556	1	0.01%
M0727	1	0.01%
M0713	1	0.01%
M0016	1	0.01%
M0349	1	0.01%
M0966	1	0.01%
M1097	1	0.01%
M0270	1	0.01%
M0211	1	0.01%
M0527	1	0.01%
M0886	1	0.01%
M0628	1	0.01%
M1071	1	0.01%
M0162	1	0.01%
M0933	1	0.01%
M0598	1	0.01%
M0433	1	0.01%
M0716	1	0.01%
M1026	1	0.01%
M0856	1	0.01%
M0970	1	0.01%
M0983	1	0.01%
M1055	1	0.01%
M0859	1	0.01%
M0688	1	0.01%
M0089	1	0.01%
M1117	1	0.01%
M0011	1	0.01%
M0506	1	0.01%
M0720	1	0.01%
M1003	1	0.01%
M1061	1	0.01%
M0414	1	0.01%
M0223	1	0.01%
M0194	1	0.01%
M0890	1	0.01%
M0161	1	0.01%
M0662	1	0.01%
M0239	1	0.01%
M0855	1	0.01%
M0303	1	0.01%
M0241	1	0.01%
M0894	1	0.01%
M0225	1	0.01%
M0617	1	0.01%
M0416	1	0.01%
M0842	1	0.01%
M0345	1	0.01%
M0913	1	0.01%
M0781	1	0.01%
M0457	1	0.01%
M0201	1	0.01%
M0552	1	0.01%
M0473	1	0.01%
M0430	1	0.01%
M0060	1	0.01%
M0528	1	0.01%
M0964	1	0.01%
M1075	1	0.01%
M0206	1	0.01%
M1120	1	0.01%
M0977	1	0.01%
M0247	1	0.01%
M1115	1	0.01%
M0995	1	0.01%
M0768	1	0.01%
M0492	1	0.01%
M1092	1	0.01%
M0900	1	0.01%
M0731	1	0.01%
M0704	1	0.01%
M0614	1	0.01%
M0575	1	0.01%
M0942	1	0.01%
M0296	1	0.01%
M0083	1	0.01%
M0553	1	0.01%
M1030	1	0.01%
M1083	1	0.01%
M0512	1	0.01%
M0500	1	0.01%
M0631	1	0.01%
M1118	1	0.01%
M1065	1	0.01%
M0214	1	0.01%
M0459	1	0.01%
M0608	1	0.01%
M0069	1	0.01%
M0412	1	0.01%
M0673	1	0.01%
M1034	1	0.01%
M0485	1	0.01%
M0880	1	0.01%
M0227	1	0.01%
M0846	1	0.01%
M0916	1	0.01%
M0373	1	0.01%
M0322	1	0.01%
M0465	1	0.01%
M0721	1	0.01%
M1009	1	0.01%
M0407	1	0.01%
M0640	1	0.01%
M0426	1	0.01%
M0577	1	0.01%
M0103	1	0.01%
M0962	1	0.01%
M0113	1	0.01%
M0334	1	0.01%
M0342	1	0.01%
M0401	1	0.01%
M0746	1	0.01%
M1018	1	0.01%
M1107	1	0.01%
M1053	1	0.01%
M0976	1	0.01%
M1038	1	0.01%
M0879	1	0.01%
M0834	1	0.01%
M1014	1	0.01%
M0562	1	0.01%
M0873	1	0.01%
M0092	1	0.01%
M0224	1	0.01%
M0994	1	0.01%
M0952	1	0.01%
M0712	1	0.01%
M1067	1	0.01%
M0599	1	0.01%
M1078	1	0.01%
M0329	1	0.01%
M1121	1	0.01%
M0691	1	0.01%
M0692	1	0.01%
M0047	1	0.01%
M0335	1	0.01%
M0979	1	0.01%
M0165	1	0.01%
M0625	1	0.01%
M0910	1	0.01%
M0602	1	0.01%
M0157	1	0.01%
M0331	1	0.01%
M0655	1	0.01%
M0097	1	0.01%
M0289	1	0.01%
M0408	1	0.01%
M0601	1	0.01%
M0667	1	0.01%
M0806	1	0.01%
M0854	1	0.01%
M0675	1	0.01%
M0730	1	0.01%
M0044	1	0.01%
M0605	1	0.01%
M0766	1	0.01%
M0698	1	0.01%
M0108	1	0.01%
M0548	1	0.01%
M0065	1	0.01%
M0561	1	0.01%
M1057	1	0.01%
M0711	1	0.01%
M0347	1	0.01%
M0078	1	0.01%
M0965	1	0.01%
M0286	1	0.01%
M0844	1	0.01%
M0586	1	0.01%
M1091	1	0.01%
M1021	1	0.01%
M0385	1	0.01%
M0782	1	0.01%
M0535	1	0.01%
M0932	1	0.01%
M0311	1	0.01%
M1051	1	0.01%
M0784	1	0.01%
M0957	1	0.01%
M0015	1	0.01%
M0516	1	0.01%
M1006	1	0.01%
M1002	1	0.01%
M1054	1	0.01%
M0432	1	0.01%
M1112	1	0.01%
M0661	1	0.01%
M0803	1	0.01%
M0293	1	0.01%
M0163	1	0.01%
M0668	1	0.01%
M0186	1	0.01%
M1088	1	0.01%
M0909	1	0.01%
M0337	1	0.01%
M0232	1	0.01%
M0937	1	0.01%
M0395	1	0.01%
M0867	1	0.01%
M0470	1	0.01%
M0950	1	0.01%
M0321	1	0.01%
M0369	1	0.01%
M0637	1	0.01%
M0265	1	0.01%
M0253	1	0.01%
M0771	1	0.01%
M1116	1	0.01%
M0425	1	0.01%
M0902	1	0.01%
M0055	1	0.01%
M0318	1	0.01%
M0340	1	0.01%
M0377	1	0.01%
M0884	1	0.01%
M0709	1	0.01%
M0574	1	0.01%
M0840	1	0.01%
M0061	1	0.01%

Value	Count	Percentage
CTY06-023	4,255	40.52%
CTY06-133	1,277	12.16%
CTY06-181	801	7.63%
CTY06-171	441	4.20%
CTY06-186	337	3.21%
CTY06-129	263	2.50%
CTY06-005	173	1.65%
CTY06-112	156	1.49%
CTY06-004	145	1.38%
CTY06-151	113	1.08%
CTY06-083	103	0.98%
CTY06-110	101	0.96%
CTY06-191	94	0.90%
CTY06-205	84	0.80%
CTY06-115	77	0.73%
CTY06-179	75	0.71%
CTY06-075	73	0.70%
CTY06-126	68	0.65%
CTY06-046	65	0.62%
CTY06-140	64	0.61%
CTY06-101	62	0.59%
CTY06-176	60	0.57%
CTY06-121	59	0.56%
CTY06-195	59	0.56%
CTY06-072	58	0.55%
CTY06-058	55	0.52%
CTY06-096	45	0.43%
CTY06-071	43	0.41%
CTY06-027	43	0.41%
CTY06-073	41	0.39%
CTY06-031	39	0.37%
CTY06-099	39	0.37%
CTY06-029	38	0.36%
CTY06-013	37	0.35%
CTY06-173	35	0.33%
CTY06-035	35	0.33%
CTY06-090	35	0.33%
CTY06-049	30	0.29%
CTY06-067	29	0.28%
CTY06-169	29	0.28%
CTY06-087	27	0.26%
CTY06-168	27	0.26%
CTY06-197	27	0.26%
CTY06-056	24	0.23%
CTY06-109	23	0.22%
CTY06-167	23	0.22%
CTY06-142	21	0.20%
CTY06-202	20	0.19%
CTY06-125	20	0.19%
CTY06-015	19	0.18%
CTY06-183	19	0.18%
CTY06-077	19	0.18%
CTY06-059	17	0.16%
CTY06-089	17	0.16%
CTY06-200	16	0.15%
CTY06-033	16	0.15%
CTY06-164	16	0.15%
CTY06-170	15	0.14%
CTY06-068	15	0.14%
CTY06-119	14	0.13%
CTY06-086	14	0.13%
CTY06-147	13	0.12%
CTY06-036	12	0.11%
CTY06-128	12	0.11%
CTY06-095	12	0.11%
CTY06-193	11	0.10%
CTY06-185	10	0.10%
CTY06-080	10	0.10%
CTY06-113	9	0.09%
CTY06-180	9	0.09%
CTY06-001	8	0.08%
CTY06-139	8	0.08%
CTY06-162	8	0.08%
CTY06-030	8	0.08%
CTY06-120	7	0.07%
CTY06-098	7	0.07%
CTY06-010	7	0.07%
CTY06-166	7	0.07%
CTY06-044	7	0.07%
CTY06-040	7	0.07%
CTY06-174	6	0.06%
CTY06-159	6	0.06%
CTY06-105	6	0.06%
CTY06-134	6	0.06%
CTY06-085	5	0.05%
CTY06-194	5	0.05%
CTY06-155	5	0.05%
CTY06-145	5	0.05%
CTY06-063	5	0.05%
CTY06-025	4	0.04%
CTY06-057	4	0.04%
CTY06-037	4	0.04%
CTY06-060	4	0.04%
CTY06-074	4	0.04%
CTY06-158	4	0.04%
CTY06-149	4	0.04%
CTY06-048	4	0.04%
CTY06-175	4	0.04%
CTY06-146	4	0.04%
CTY06-002	4	0.04%
CTY06-022	4	0.04%
CTY06-064	4	0.04%
CTY06-051	3	0.03%
CTY06-154	3	0.03%
CTY06-192	3	0.03%
CTY06-152	3	0.03%
CTY06-117	3	0.03%
CTY06-034	3	0.03%
CTY07-002	3	0.03%
CTY06-093	3	0.03%
CTY06-039	3	0.03%
CTY06-123	3	0.03%
CTY06-009	3	0.03%
CTY06-021	3	0.03%
CTY06-104	3	0.03%
CTY06-079	2	0.02%
CTY06-062	2	0.02%
CTY05-005	2	0.02%
CTY06-189	2	0.02%
CTY06-061	2	0.02%
CTY06-102	2	0.02%
CTY06-003	2	0.02%
CTY06-172	2	0.02%
CTY06-108	2	0.02%
CTY06-182	2	0.02%
CTY06-082	2	0.02%
CTY06-028	2	0.02%
CTY06-006	2	0.02%
CTY06-160	2	0.02%
CTY06-018	2	0.02%
CTY08-002	2	0.02%
CTY06-203	2	0.02%
CTY06-012	2	0.02%
CTY05-004	2	0.02%
CTY06-017	2	0.02%
CTY06-020	2	0.02%
CTY06-131	2	0.02%
CTY06-007	2	0.02%
CTY06-153	2	0.02%
CTY06-094	2	0.02%
CTY06-103	2	0.02%
CTY06-038	2	0.02%
CTY06-092	2	0.02%
CTY06-190	2	0.02%
CTY06-111	2	0.02%
CTY06-042	2	0.02%
CTY13-002	1	0.01%
CTY06-008	1	0.01%
CTY08-006	1	0.01%
CTY06-019	1	0.01%
CTY06-114	1	0.01%
CTY06-116	1	0.01%
CTY10-001	1	0.01%
CTY06-011	1	0.01%
CTY07-003	1	0.01%
CTY13-003	1	0.01%
CTY09-002	1	0.01%
CTY06-137	1	0.01%
CTY06-148	1	0.01%
CTY06-100	1	0.01%
CTY06-014	1	0.01%
CTY05-003	1	0.01%
CTY06-201	1	0.01%
CTY02-001	1	0.01%
CTY12-002	1	0.01%
CTY06-052	1	0.01%
CTY08-004	1	0.01%
CTY06-078	1	0.01%
CTY06-076	1	0.01%
CTY06-097	1	0.01%
CTY06-204	1	0.01%
CTY01-002	1	0.01%
CTY05-002	1	0.01%
CTY10-002	1	0.01%
CTY06-130	1	0.01%
CTY06-065	1	0.01%
CTY06-127	1	0.01%
CTY06-165	1	0.01%
CTY06-199	1	0.01%
CTY08-005	1	0.01%
CTY06-150	1	0.01%
CTY06-163	1	0.01%
CTY06-144	1	0.01%
CTY04-001	1	0.01%
CTY06-084	1	0.01%
CTY06-088	1	0.01%
CTY08-007	1	0.01%
CTY06-047	1	0.01%
CTY06-177	1	0.01%
CTY06-122	1	0.01%
CTY06-069	1	0.01%
CTY06-043	1	0.01%
CTY07-001	1	0.01%
CTY12-001	1	0.01%
CTY08-003	1	0.01%
CTY01-001	1	0.01%
CTY06-106	1	0.01%
CTY08-001	1	0.01%
CTY06-124	1	0.01%
CTY05-001	1	0.01%
CTY06-050	1	0.01%
CTY06-143	1	0.01%
CTY06-026	1	0.01%
CTY01-003	1	0.01%
CTY06-053	1	0.01%
CTY05-007	1	0.01%
CTY06-136	1	0.01%
CTY06-188	1	0.01%
CTY06-157	1	0.01%
CTY06-198	1	0.01%
CTY06-024	1	0.01%
CTY06-141	1	0.01%
CTY06-161	1	0.01%
CTY13-001	1	0.01%
CTY06-178	1	0.01%
CTY06-138	1	0.01%
CTY06-132	1	0.01%
CTY06-107	1	0.01%
CTY06-135	1	0.01%
CTY08-008	1	0.01%
CTY09-001	1	0.01%
CTY06-187	1	0.01%
CTY06-184	1	0.01%
CTY06-041	1	0.01%
CTY06-091	1	0.01%
CTY06-054	1	0.01%
CTY06-118	1	0.01%

Fraud_Status	0	1
Month
0	126	12
1	52	3
2	122	17
3	600	50
4	1074	72
5	1280	84
6	1219	78
7	1176	79
8	1214	68
9	1180	57
10	797	44
11	308	21

	Variable 1	Variable 2	PPS 1	PPS 2	Correlation	P-Value
1	Channel_ID	Merchant_ID	36.34%	41.19%	78.93%	0.00%
2	Maximum_Transaction_Amount	Average_Transaction_Amount	45.16%	39.10%	69.23%	0.00%
3	Channel_ID	Card_Holder	0.00%	0.00%	52.57%	0.00%
4	Card_Type	Maximum_Transaction_Amount	0.00%	10.69%	46.85%	0.00%
5	Transaction_Amount	Average_Transaction_Amount	5.40%	0.00%	45.53%	0.00%
6	Card_Type	Average_Transaction_Amount	0.52%	66.03%	32.88%	0.00%
7	Average_Transaction_Frequency	Maximum_Transaction_Amount	13.26%	3.76%	30.74%	0.00%
8	Transaction_Amount	Maximum_Transaction_Amount	0.00%	0.00%	27.89%	0.00%
9	Minimum_Transaction_Amount	Average_Transaction_Amount	14.73%	0.00%	24.82%	0.00%
10	Average_Transaction_Frequency	Average_Transaction_Amount	9.95%	0.00%	21.39%	0.00%
11	Average_Transaction_Frequency	Card_Type	0.00%	0.00%	20.10%	0.00%
12	Transaction_Amount	Card_Type	0.00%	0.00%	15.43%	0.00%
13	Transaction_Amount	Minimum_Transaction_Amount	0.00%	0.00%	14.00%	0.00%
14	City_ID	Card_Holder	0.00%	0.00%	-11.88%	0.00%
15	Card_Holder	Merchant_ID	0.00%	0.00%	-10.74%	0.00%
16	Transaction_Amount	Merchant_ID	10.25%	0.00%	10.56%	0.00%
17	Transaction_Amount	Average_Transaction_Frequency	0.00%	0.00%	10.54%	0.00%
18	Transaction_Amount	Channel_ID	0.00%	0.00%	9.81%	0.00%
19	Minimum_Transaction_Amount	Merchant_ID	0.00%	0.00%	6.82%	0.00%
20	Average_Transaction_Amount	Merchant_ID	0.00%	0.00%	6.65%	0.00%
21	Maximum_Transaction_Amount	Minimum_Transaction_Amount	0.00%	2.57%	5.65%	0.00%
22	City_ID	Channel_ID	0.00%	0.00%	-5.63%	0.00%
23	Channel_ID	Average_Transaction_Amount	0.00%	0.00%	5.31%	0.00%
24	Average_Transaction_Frequency	Minimum_Transaction_Amount	0.00%	0.00%	-4.56%	0.00%
25	Card_Holder	Minimum_Transaction_Amount	0.00%	0.00%	-4.12%	0.00%
26	Card_Type	Minimum_Transaction_Amount	0.00%	0.00%	3.58%	0.04%
27	Channel_ID	Minimum_Transaction_Amount	0.00%	0.00%	3.29%	0.12%
28	Channel_ID	Maximum_Transaction_Amount	0.00%	0.00%	2.55%	1.20%
29	Average_Transaction_Frequency	Channel_ID	0.00%	0.00%	2.54%	1.23%
30	Average_Transaction_Frequency	Merchant_ID	0.00%	0.00%	2.48%	1.43%
31	Maximum_Transaction_Amount	Merchant_ID	0.00%	0.00%	2.37%	1.95%
32	City_ID	Minimum_Transaction_Amount	0.00%	0.00%	2.23%	2.80%
33	City_ID	Merchant_ID	0.00%	0.00%	2.00%	4.84%
34	City_ID	Card_Type	0.00%	0.00%	1.90%	6.05%
35	Transaction_Amount	City_ID	0.00%	0.00%	1.50%	13.81%
36	Transaction_Amount	Card_Holder	0.00%	0.00%	1.25%	21.86%
37	Card_Holder	Maximum_Transaction_Amount	0.00%	0.00%	0.84%	40.63%
38	Average_Transaction_Frequency	Card_Holder	0.00%	0.00%	0.67%	50.88%
39	Card_Type	Channel_ID	0.00%	0.00%	0.62%	54.10%
40	Card_Holder	Average_Transaction_Amount	0.00%	0.00%	-0.61%	54.67%
41	Card_Type	Merchant_ID	0.00%	0.00%	0.60%	55.43%
42	City_ID	Average_Transaction_Amount	0.00%	0.00%	0.41%	68.59%
43	Average_Transaction_Frequency	City_ID	0.00%	0.00%	-0.39%	70.03%
44	City_ID	Maximum_Transaction_Amount	0.00%	0.00%	-0.32%	75.55%
45	Card_Type	Card_Holder	0.00%	0.00%	0.17%	86.47%

H2O_cluster_uptime:	17 days 11 hours 3 mins
H2O_cluster_timezone:	Asia/Jakarta
H2O_data_parsing_timezone:	UTC
H2O_cluster_version:	3.32.1.3
H2O_cluster_version_age:	1 month and 26 days
H2O_cluster_name:	H2O_from_python_charissa_9lw8gd
H2O_cluster_total_nodes:	1
H2O_cluster_free_memory:	1.159 Gb
H2O_cluster_total_cores:	8
H2O_cluster_allowed_cores:	8
H2O_cluster_status:	locked, healthy
H2O_connection_url:	http://localhost:54321
H2O_connection_proxy:	{"http": null, "https": null}
H2O_internal_security:	False
H2O_API_Extensions:	Amazon S3, XGBoost, Algos, AutoML, Core V3, TargetEncoder, Core V4
Python_version:	3.8.8 final

		0	1	Error	Rate
0	0	7036.0	282.0	0.0385	(282.0/7318.0)
1	1	271.0	197.0	0.5791	(271.0/468.0)
2	Total	7307.0	479.0	0.071	(553.0/7786.0)

	metric	threshold	value	idx
0	max f1	0.295121	0.416051	139.0
1	max f2	0.063268	0.450928	227.0
2	max f0point5	0.364045	0.432135	97.0
3	max accuracy	0.447983	0.941947	40.0
4	max precision	0.838133	1.000000	0.0
5	max recall	0.017840	1.000000	380.0
6	max specificity	0.838133	1.000000	0.0
7	max absolute_mcc	0.295121	0.378273	139.0
8	max min_per_class_accuracy	0.040735	0.698718	284.0
9	max mean_per_class_accuracy	0.052311	0.724485	246.0
10	max tns	0.838133	7318.000000	0.0
11	max fns	0.838133	467.000000	0.0
12	max fps	0.001284	7318.000000	399.0
13	max tps	0.017840	468.000000	380.0
14	max tnr	0.838133	1.000000	0.0
15	max fnr	0.838133	0.997863	0.0
16	max fpr	0.001284	1.000000	399.0
17	max tpr	0.017840	1.000000	380.0

Value	Count	Percentage
0	9,772	93.07%
1	728	6.93%

Value	Count	Percentage
1	8,325	79.29%
2	1,463	13.93%
5	682	6.50%
3	25	0.24%
4	5	0.05%

	Card_Type	Fraud_Status
0	0	0.064023
1	1	0.045117

	Card_Holder	Fraud_Status
0	0	0.381538
1	1	0.037102

	group	cumulative_data_fraction	lower_threshold	lift	cumulative_lift	response_rate	score	cumulative_response_rate	cumulative_score	capture_rate	cumulative_capture_rate	gain	cumulative_gain	kolmogorov_smirnov
0	1	0.010018	0.438414	9.598126	9.598126	0.576923	0.508515	0.576923	0.508515	0.096154	0.096154	859.812623	859.812623	0.091644
1	2	0.020036	0.408233	7.465209	8.531668	0.448718	0.421909	0.512821	0.465212	0.074786	0.170940	646.520929	753.166776	0.160555
2	3	0.030054	0.381625	7.465209	8.176182	0.448718	0.394570	0.491453	0.441665	0.074786	0.245726	646.520929	717.618161	0.229465
3	4	0.040072	0.364233	6.825334	7.838470	0.410256	0.373182	0.471154	0.424544	0.068376	0.314103	582.533421	683.846976	0.291555
4	5	0.050090	0.337932	4.905709	7.251918	0.294872	0.351083	0.435897	0.409852	0.049145	0.363248	390.570896	625.191760	0.333185
5	6	0.100051	0.062547	2.950992	5.104215	0.177378	0.168332	0.306804	0.289247	0.147436	0.510684	195.099202	410.421535	0.436893
6	7	0.150013	0.049639	1.154736	3.788849	0.069409	0.054386	0.227740	0.211027	0.057692	0.568376	15.473601	278.884937	0.445118
7	8	0.200103	0.045639	0.981142	3.086021	0.058974	0.047469	0.185494	0.170085	0.049145	0.617521	-1.885821	208.602142	0.444113
8	9	0.300026	0.041304	0.705672	2.293251	0.042416	0.043220	0.137842	0.127833	0.070513	0.688034	-29.432799	129.325094	0.412822
9	10	0.400077	0.038321	0.405774	1.821230	0.024390	0.039729	0.109470	0.105800	0.040598	0.728632	-59.422556	82.123033	0.349567
10	11	0.500000	0.035868	0.769824	1.611111	0.046272	0.037121	0.096840	0.092075	0.076923	0.805556	-23.017599	61.111111	0.325096
11	12	0.600051	0.033556	0.619340	1.445745	0.037227	0.034712	0.086901	0.082510	0.061966	0.867521	-38.066006	44.574516	0.284575
12	13	0.699974	0.030982	0.406296	1.297361	0.024422	0.032316	0.077982	0.075345	0.040598	0.908120	-59.370400	29.736141	0.221457
13	14	0.800026	0.027909	0.341705	1.177847	0.020539	0.029519	0.070798	0.069614	0.034188	0.942308	-65.829521	17.784680	0.151381
14	15	0.899949	0.023118	0.320760	1.082683	0.019280	0.025774	0.065078	0.064746	0.032051	0.974359	-67.924000	8.268288	0.079169
15	16	1.000000	0.001099	0.256279	1.000000	0.015404	0.018387	0.060108	0.060108	0.025641	1.000000	-74.372140	0.000000	0.000000

		0	1	Error	Rate
0	0	1747.0	83.0	0.0454	(83.0/1830.0)
1	1	69.0	48.0	0.5897	(69.0/117.0)
2	Total	1816.0	131.0	0.0781	(152.0/1947.0)

	metric	threshold	value	idx
0	max f1	0.108795	0.387097	114.0
1	max f2	0.108795	0.400668	114.0
2	max f0point5	0.260051	0.381282	98.0
3	max accuracy	0.575565	0.940421	2.0
4	max precision	0.575565	0.666667	2.0
5	max recall	0.019202	1.000000	369.0
6	max specificity	0.811500	0.999454	0.0
7	max absolute_mcc	0.108795	0.346179	114.0
8	max min_per_class_accuracy	0.040037	0.662842	237.0
9	max mean_per_class_accuracy	0.057428	0.685785	156.0
10	max tns	0.811500	1829.000000	0.0
11	max fns	0.811500	117.000000	0.0
12	max fps	0.001316	1830.000000	399.0
13	max tps	0.019202	117.000000	369.0
14	max tnr	0.811500	0.999454	0.0
15	max fnr	0.811500	1.000000	0.0
16	max fpr	0.001316	1.000000	399.0
17	max tpr	0.019202	1.000000	369.0

	group	cumulative_data_fraction	lower_threshold	lift	cumulative_lift	response_rate	score	cumulative_response_rate	cumulative_score	capture_rate	cumulative_capture_rate	gain	cumulative_gain	kolmogorov_smirnov
0	1	0.010272	0.430373	4.992308	4.992308	0.300000	0.502404	0.300000	0.502404	0.051282	0.051282	399.230769	399.230769	0.043632
1	2	0.020031	0.395597	7.882591	6.400394	0.473684	0.411902	0.384615	0.458314	0.076923	0.128205	688.259109	540.039448	0.115090
2	3	0.030303	0.379552	5.824359	6.205128	0.350000	0.387677	0.372881	0.434369	0.059829	0.188034	482.435897	520.512821	0.167816
3	4	0.040062	0.355861	6.130904	6.187048	0.368421	0.366353	0.371795	0.417801	0.059829	0.247863	513.090418	518.704799	0.221087
4	5	0.050334	0.311663	7.488462	6.452643	0.450000	0.338321	0.387755	0.401581	0.076923	0.324786	648.846154	545.264260	0.291999
5	6	0.100154	0.057917	2.401798	4.437607	0.144330	0.130577	0.266667	0.266774	0.119658	0.444444	140.179752	343.760684	0.366302
6	7	0.149974	0.048796	0.343114	3.077450	0.020619	0.053267	0.184932	0.195848	0.017094	0.461538	-65.688607	207.744995	0.331484
7	8	0.200308	0.045096	0.849032	2.517488	0.051020	0.046789	0.151282	0.158392	0.042735	0.504274	-15.096808	151.748849	0.323399
8	9	0.299949	0.041602	1.029342	2.023138	0.061856	0.043105	0.121575	0.120095	0.102564	0.606838	2.934179	102.313839	0.326510
9	10	0.400103	0.038567	1.280079	1.837135	0.076923	0.040124	0.110398	0.100077	0.128205	0.735043	28.007890	83.713505	0.356354
10	11	0.500257	0.036109	0.512032	1.571842	0.030769	0.037345	0.094456	0.087517	0.051282	0.786325	-48.796844	57.184226	0.304358
11	12	0.599897	0.033822	0.686228	1.424745	0.041237	0.034981	0.085616	0.078791	0.068376	0.854701	-31.377214	42.474535	0.271094
12	13	0.700051	0.031282	0.768047	1.330794	0.046154	0.032587	0.079971	0.072181	0.076923	0.931624	-23.195266	33.079369	0.246378
13	14	0.799692	0.027963	0.171557	1.186354	0.010309	0.029675	0.071291	0.066885	0.017094	0.948718	-82.844303	18.635443	0.158554
14	15	0.899846	0.022994	0.256016	1.082806	0.015385	0.025738	0.065068	0.062305	0.025641	0.974359	-74.398422	8.280646	0.079277
15	16	1.000000	0.001316	0.256016	1.000000	0.015385	0.018793	0.060092	0.057947	0.025641	1.000000	-74.398422	0.000000	0.000000

		0	1	Error	Rate
0	0	6992.0	326.0	0.0445	(326.0/7318.0)
1	1	258.0	210.0	0.5513	(258.0/468.0)
2	Total	7250.0	536.0	0.075	(584.0/7786.0)

	metric	threshold	value	idx
0	max f1	0.119526	0.418327	190.0
1	max f2	0.062429	0.447217	227.0
2	max f0point5	0.371643	0.433468	91.0
3	max accuracy	0.450977	0.941819	39.0
4	max precision	0.844948	1.000000	0.0
5	max recall	0.015479	1.000000	386.0
6	max specificity	0.844948	1.000000	0.0
7	max absolute_mcc	0.282622	0.380178	144.0
8	max min_per_class_accuracy	0.041906	0.692308	279.0
9	max mean_per_class_accuracy	0.062429	0.714769	227.0
10	max tns	0.844948	7318.000000	0.0
11	max fns	0.844948	467.000000	0.0
12	max fps	0.001381	7318.000000	399.0
13	max tps	0.015479	468.000000	386.0
14	max tnr	0.844948	1.000000	0.0
15	max fnr	0.844948	0.997863	0.0
16	max fpr	0.001381	1.000000	399.0
17	max tpr	0.015479	1.000000	386.0

	group	cumulative_data_fraction	lower_threshold	lift	cumulative_lift	response_rate	score	cumulative_response_rate	cumulative_score	capture_rate	cumulative_capture_rate	gain	cumulative_gain	kolmogorov_smirnov
0	1	0.010018	0.435563	9.171543	9.171543	0.551282	0.505493	0.551282	0.505493	0.091880	0.091880	817.154284	817.154284	0.087098
1	2	0.020036	0.398264	7.678501	8.425022	0.461538	0.414682	0.506410	0.460088	0.076923	0.168803	667.850099	742.502192	0.158281
2	3	0.030054	0.375837	8.105084	8.318376	0.487179	0.386777	0.500000	0.435651	0.081197	0.250000	710.508437	731.837607	0.234012
3	4	0.040072	0.355259	5.758876	7.678501	0.346154	0.366244	0.461538	0.418299	0.057692	0.307692	475.887574	667.850099	0.284735
4	5	0.050090	0.322117	5.758876	7.294576	0.346154	0.339368	0.438462	0.402513	0.057692	0.365385	475.887574	629.457594	0.335458
5	6	0.129335	0.061129	1.887476	3.981586	0.113452	0.111129	0.239325	0.223979	0.149573	0.514957	88.747593	298.158616	0.410284
6	7	0.162856	0.060856	0.637423	3.293237	0.038314	0.060856	0.197950	0.190402	0.021368	0.536325	-36.257655	229.323721	0.397352
7	8	0.200103	0.050559	1.204730	2.904491	0.072414	0.054781	0.174583	0.165158	0.044872	0.581197	20.473033	190.449075	0.405466
8	9	0.300026	0.043055	0.855360	2.222032	0.051414	0.046095	0.133562	0.125505	0.085470	0.666667	-14.463999	122.203196	0.390088
9	10	0.400077	0.039443	0.576627	1.810549	0.034660	0.041168	0.108828	0.104414	0.057692	0.724359	-42.337316	81.054863	0.345020
10	11	0.500000	0.036764	0.555984	1.559829	0.033419	0.038073	0.093758	0.091156	0.055556	0.779915	-44.401600	55.982906	0.297816
11	12	0.600051	0.034249	0.619340	1.403014	0.037227	0.035493	0.084332	0.081875	0.061966	0.841880	-38.066006	40.301377	0.257294
12	13	0.699974	0.031651	0.598752	1.288204	0.035990	0.032966	0.077431	0.074893	0.059829	0.901709	-40.124800	28.820356	0.214636
13	14	0.800026	0.028524	0.277635	1.161822	0.016688	0.030125	0.069835	0.069294	0.027778	0.929487	-72.236486	16.182167	0.137741
14	15	0.899949	0.023577	0.363528	1.073186	0.021851	0.026329	0.064507	0.064524	0.036325	0.965812	-63.647200	7.318567	0.070075
15	16	1.000000	0.000854	0.341705	1.000000	0.020539	0.018776	0.060108	0.059947	0.034188	1.000000	-65.829521	0.000000	0.000000

		mean	sd	cv_1_valid	cv_2_valid	cv_3_valid	cv_4_valid	cv_5_valid	cv_6_valid	cv_7_valid	cv_8_valid	cv_9_valid	cv_10_valid	cv_11_valid	cv_12_valid	cv_13_valid	cv_14_valid	cv_15_valid	cv_16_valid	cv_17_valid	cv_18_valid	cv_19_valid	cv_20_valid	cv_21_valid	cv_22_valid	cv_23_valid	cv_24_valid	cv_25_valid	cv_26_valid	cv_27_valid	cv_28_valid	cv_29_valid	cv_30_valid
0	accuracy	0.9319444	0.02676247	0.9346154	0.9461538	0.95	0.8346154	0.91923076	0.9269231	0.91923076	0.93846154	0.89615387	0.9461538	0.9115385	0.9346154	0.93846154	0.96153843	0.9115385	0.9230769	0.94208497	0.9150579	0.9266409	0.96525097	0.96138996	0.9150579	0.93822396	0.95752895	0.9227799	0.93822396	0.9111969	0.976834	0.93050194	0.96525097
1	auc	0.7699643	0.08766349	0.77792007	0.7270748	0.574	0.7502083	0.80246913	0.8624154	0.68395835	0.66071427	0.8169935	0.8262167	0.5684	0.79297173	0.55188245	0.8780952	0.75	0.79602534	0.83577406	0.8	0.7889121	0.8765432	0.74890554	0.7942623	0.86938775	0.88742965	0.7528815	0.8031381	0.7568306	0.7549801	0.8197017	0.79083663
2	err	0.06805564	0.02676247	0.06538462	0.053846154	0.05	0.16538462	0.08076923	0.073076926	0.08076923	0.06153846	0.103846155	0.053846154	0.08846154	0.06538462	0.06153846	0.03846154	0.08846154	0.07692308	0.057915058	0.08494209	0.07335907	0.034749035	0.038610037	0.08494209	0.06177606	0.042471044	0.077220075	0.06177606	0.08880309	0.023166023	0.06949807	0.034749035
3	err_count	17.666666	6.959654	17.0	14.0	13.0	43.0	21.0	19.0	21.0	16.0	27.0	14.0	23.0	17.0	16.0	10.0	23.0	20.0	15.0	22.0	19.0	9.0	10.0	22.0	16.0	11.0	20.0	16.0	23.0	6.0	18.0	9.0
4	f0point5	0.46692804	0.13924657	0.46052632	0.49019608	0.32608697	0.28846154	0.45454547	0.5217391	0.41666666	0.234375	0.3649635	0.6060606	0.1744186	0.6097561	0.5	0.67164177	0.2777778	0.443038	0.625	0.34313726	0.52884614	0.7692308	0.6122449	0.40650406	0.4651163	0.5660377	0.45918366	0.5952381	0.36036035	0.625	0.45454547	0.35714287
5	f1	0.45632124	0.110560715	0.4516129	0.41666666	0.31578946	0.35820895	0.5116279	0.55813956	0.36363637	0.27272728	0.42553192	0.53333336	0.20689656	0.5405405	0.3846154	0.64285713	0.3030303	0.4117647	0.5945946	0.3888889	0.5365854	0.64	0.54545456	0.47619048	0.5	0.5217391	0.47368422	0.5555556	0.41025642	0.5714286	0.47058824	0.30769232
6	f2	0.45867836	0.10390178	0.443038	0.36231884	0.30612245	0.47244096	0.5851064	0.6	0.32258064	0.32608697	0.5102041	0.47619048	0.2542373	0.4854369	0.3125	0.6164383	0.33333334	0.3846154	0.5670103	0.44871795	0.5445545	0.5479452	0.4918033	0.57471263	0.5405405	0.48387095	0.48913044	0.5208333	0.47619048	0.5263158	0.4878049	0.27027026
7	lift_top_group	9.01935	3.257432	10.833333	11.555555	8.666667	8.666667	10.196078	9.122807	8.666667	0.0	15.294118	9.62963	8.666667	11.818182	9.62963	5.7777777	6.1904764	9.122807	12.95	6.1666665	8.633333	16.1875	13.282051	5.7555556	6.1666665	6.6410255	9.592592	8.633333	5.7555556	10.791667	5.3958335	10.791667
8	logloss	0.18337043	0.032842204	0.1876225	0.1853218	0.15685907	0.24966325	0.1874058	0.18505175	0.246197	0.13779983	0.19819152	0.1874985	0.1609587	0.22206269	0.22799125	0.14168519	0.19228448	0.22527215	0.19010212	0.18087101	0.19930607	0.15932454	0.15547653	0.17896676	0.15689974	0.14345263	0.20933427	0.20883945	0.18930222	0.12129969	0.18843003	0.1276423
9	max_per_class_error	0.5313485	0.11762855	0.5625	0.6666667	0.7	0.4	0.3529412	0.36842105	0.7	0.625	0.4117647	0.5555556	0.7	0.54545456	0.7222222	0.4	0.64285713	0.6315789	0.45	0.5	0.45	0.5	0.53846157	0.33333334	0.42857143	0.53846157	0.5	0.5	0.46666667	0.5	0.5	0.75
10	mcc	0.4313766	0.11506222	0.417132	0.40433297	0.29034942	0.314483	0.4823187	0.52310306	0.3311331	0.25351265	0.39146608	0.5177088	0.17438	0.51748455	0.39009342	0.6244224	0.26037762	0.3741919	0.5657891	0.3558803	0.49695313	0.65185744	0.5357057	0.45633632	0.47182348	0.5046277	0.43284053	0.5265611	0.3768111	0.5658994	0.43439713	0.29932582
11	mean_per_class_accuracy	0.7147494	0.054424644	0.7023566	0.6585034	0.638	0.7270833	0.79266524	0.7908932	0.6354167	0.6656746	0.7529654	0.7139578	0.618	0.7167685	0.63269055	0.79183674	0.65011615	0.667613	0.7624477	0.71938777	0.7540795	0.7479424	0.72467166	0.79849726	0.7653061	0.72263914	0.7271784	0.7374477	0.7338798	0.74601597	0.7294239	0.6190239
12	mean_per_class_error	0.2852506	0.054424644	0.29764345	0.3414966	0.362	0.27291667	0.20733479	0.20910679	0.36458334	0.3343254	0.24703461	0.28604224	0.382	0.28323147	0.36730945	0.20816326	0.34988385	0.33238697	0.2375523	0.28061223	0.24592051	0.2520576	0.27532834	0.20150273	0.23469388	0.27736086	0.27282158	0.2625523	0.26612023	0.25398406	0.27057612	0.38097608
13	mse	0.047172975	0.010062259	0.04888559	0.046554737	0.036305137	0.06667076	0.049683847	0.0510059	0.06478831	0.030207302	0.052116413	0.048295792	0.036631126	0.058904957	0.058934156	0.036906075	0.048691798	0.05977394	0.049625657	0.047268216	0.053533044	0.040459316	0.03701342	0.048762523	0.041314267	0.03638495	0.05464109	0.055399504	0.048425395	0.02751335	0.049752545	0.030740079
14	null_deviance	118.03537	21.421686	120.223526	114.7255	87.37807	142.31174	125.73113	136.7752	142.31174	76.50513	125.73113	131.24835	87.37807	153.41394	131.24835	114.7255	109.23703	136.7752	142.19029	109.11213	142.19029	120.09978	103.63261	114.60118	109.11213	103.63261	131.12575	142.19029	114.60118	76.37677	120.09978	76.37677
15	pr_auc	0.3289946	0.14216518	0.30546305	0.34000897	0.107951894	0.25836605	0.40915602	0.45839494	0.2970723	0.08474422	0.41740307	0.48763844	0.083102316	0.52515036	0.24085344	0.5193918	0.15094063	0.2903547	0.5837953	0.21639653	0.42677876	0.61288583	0.3958843	0.23024206	0.27695373	0.405081	0.3399414	0.4308292	0.27877495	0.2712737	0.26067606	0.16433305
16	precision	0.48201114	0.16940746	0.46666667	0.5555556	0.33333334	0.25531915	0.42307693	0.5	0.46153846	0.21428572	0.33333334	0.6666667	0.15789473	0.6666667	0.625	0.6923077	0.2631579	0.46666667	0.64705884	0.3181818	0.52380955	0.8888889	0.6666667	0.37037036	0.44444445	0.6	0.45	0.625	0.33333334	0.6666667	0.44444445	0.4
17	r2	0.14914195	0.09675528	0.15351795	0.14364623	0.018309088	0.061053462	0.18696974	0.2469974	0.08756466	-0.012903579	0.14716302	0.25050607	0.009494395	0.23950055	0.08541116	0.32112905	0.044260897	0.1175544	0.30356932	0.075568765	0.24873431	0.3019415	0.22360942	0.10627356	0.19201101	0.23679207	0.15505323	0.22254097	0.11245246	0.08086506	0.14160222	-0.026929885
18	recall	0.46865144	0.11762855	0.4375	0.33333334	0.3	0.6	0.64705884	0.6315789	0.3	0.375	0.5882353	0.44444445	0.3	0.45454547	0.2777778	0.6	0.35714287	0.36842105	0.55	0.5	0.55	0.5	0.46153846	0.6666667	0.5714286	0.46153846	0.5	0.5	0.53333336	0.5	0.5	0.25
19	residual_deviance	95.19201	17.113794	97.5637	96.36733	81.56671	129.82489	97.45101	96.22691	128.02245	71.655914	103.05959	97.499214	83.698524	115.4726	118.55546	73.6763	99.98792	117.14152	98.47289	93.691185	103.24055	82.53011	80.536835	92.70478	81.27406	74.30846	108.43515	108.17883	98.05855	62.83324	97.60676	66.11871

	timestamp	duration	iteration	lambda	predictors	deviance_train	deviance_test	deviance_xval	deviance_se	iterations	training_rmse	training_logloss	training_r2	training_auc	training_pr_auc	training_lift	training_classification_error	validation_rmse	validation_logloss	validation_r2	validation_auc	validation_pr_auc	validation_lift	validation_classification_error
0	2021-07-15 20:11:18	0.000 sec	2	.89E1	15	0.452069	0.452454	0.452427	0.014922	NaN
1	2021-07-15 20:11:18	0.003 sec	4	.56E1	15	0.450602	0.451268	0.451025	0.014849	NaN
2	2021-07-15 20:11:18	0.005 sec	6	.34E1	15	0.448294	0.449405	0.448817	0.014734	NaN
3	2021-07-15 20:11:18	0.008 sec	8	.21E1	15	0.444701	0.446508	0.445375	0.014556	NaN
4	2021-07-15 20:11:18	0.011 sec	10	.13E1	15	0.439270	0.442140	0.440160	0.014291	NaN
5	2021-07-15 20:11:18	0.013 sec	12	.83E0	15	0.431409	0.435843	0.432580	0.013915	NaN
6	2021-07-15 20:11:18	0.016 sec	14	.51E0	15	0.420860	0.427445	0.422346	0.013425	NaN
7	2021-07-15 20:11:18	0.018 sec	16	.32E0	15	0.408390	0.417623	0.410127	0.012877	NaN
8	2021-07-15 20:11:18	0.021 sec	18	.2E0	15	0.396009	0.408041	0.397861	0.012385	NaN
9	2021-07-15 20:11:18	0.023 sec	20	.12E0	15	0.385851	0.400388	0.387739	0.012047	NaN
10	2021-07-15 20:11:18	0.026 sec	22	.76E-1	15	0.378718	0.395230	0.380675	0.011877	NaN
11	2021-07-15 20:11:18	0.029 sec	24	.47E-1	15	0.374173	0.392135	0.376288	0.011828	NaN
12	2021-07-15 20:11:18	0.031 sec	26	.29E-1	15	0.371398	0.390418	0.373748	0.011848	NaN
13	2021-07-15 20:11:18	0.034 sec	28	.18E-1	15	0.369710	0.389525	0.372337	0.011901	NaN
14	2021-07-15 20:11:18	0.039 sec	30	.11E-1	15	0.368679	0.389112	0.371551	0.011971	30.0	0.217996	0.184026	0.158821	0.779402	0.31251	9.598126	0.071025	0.224215	0.194489	0.109931	0.745708	0.221553	4.992308	0.078069
15	2021-07-15 20:11:18	0.042 sec	32	.7E-2	15	0.368053	0.388979	0.371333	0.011990	NaN
16	2021-07-15 20:11:18	0.044 sec	34	.44E-2	15	0.367681	0.389006	0.373396	0.012335	NaN
17	2021-07-15 20:11:18	0.050 sec	36	.27E-2	15	0.367466	0.389112	0.374819	0.013044	NaN
18	2021-07-15 20:11:18	0.052 sec	37	.17E-2	15	0.367338	0.389250	0.376569	0.013254	NaN

	variable	relative_importance	scaled_importance	percentage
0	Card_Holder	0.578155	1.000000	0.271839
1	Average_Transaction_Frequency	0.258025	0.446290	0.121319
2	Merchant_ID	0.208916	0.361349	0.098229
3	Card_Type.1	0.201374	0.348304	0.094683
4	Card_Type.0	0.199199	0.344543	0.093660
5	Channel_ID	0.179380	0.310262	0.084341
6	Minimum_Transaction_Amount	0.169619	0.293380	0.079752
7	Transaction_Amount	0.107500	0.185937	0.050545
8	Transaction_Date	0.080574	0.139363	0.037884
9	Average_Transaction_Amount	0.057392	0.099268	0.026985
10	Day	0.030577	0.052888	0.014377
11	Maximum_Transaction_Amount	0.025478	0.044067	0.011979
12	Month	0.016241	0.028092	0.007636
13	City_ID	0.014397	0.024902	0.006769

		0	1	Error	Rate
0	0	7012.0	306.0	0.0418	(306.0/7318.0)
1	1	274.0	194.0	0.5855	(274.0/468.0)
2	Total	7286.0	500.0	0.0745	(580.0/7786.0)

	metric	threshold	value	idx
0	max f1	0.254164	0.400826	149.0
1	max f2	0.066169	0.428465	230.0
2	max f0point5	0.305003	0.407088	123.0
3	max accuracy	0.442060	0.940406	40.0
4	max precision	0.827848	1.000000	0.0
5	max recall	0.017389	1.000000	383.0
6	max specificity	0.827848	1.000000	0.0
7	max absolute_mcc	0.254164	0.361382	149.0
8	max min_per_class_accuracy	0.042450	0.681624	286.0
9	max mean_per_class_accuracy	0.063060	0.707160	234.0
10	max tns	0.827848	7318.000000	0.0
11	max fns	0.827848	467.000000	0.0
12	max fps	0.000938	7318.000000	399.0
13	max tps	0.017389	468.000000	383.0
14	max tnr	0.827848	1.000000	0.0
15	max fnr	0.827848	0.997863	0.0
16	max fpr	0.000938	1.000000	399.0
17	max tpr	0.017389	1.000000	383.0

	group	cumulative_data_fraction	lower_threshold	lift	cumulative_lift	response_rate	score	cumulative_response_rate	cumulative_score	capture_rate	cumulative_capture_rate	gain	cumulative_gain	kolmogorov_smirnov
0	1	0.010018	0.425251	8.318376	8.318376	0.500000	0.504983	0.500000	0.504983	0.083333	0.083333	731.837607	731.837607	0.078004
1	2	0.020036	0.385907	7.465209	7.891793	0.448718	0.402645	0.474359	0.453814	0.074786	0.158120	646.520929	689.179268	0.146914
2	3	0.030054	0.356530	7.251918	7.678501	0.435897	0.370036	0.461538	0.425888	0.072650	0.230769	625.191760	667.850099	0.213551
3	4	0.040072	0.336618	6.398751	7.358563	0.384615	0.345287	0.442308	0.405738	0.064103	0.294872	539.875082	635.856345	0.271095
4	5	0.050090	0.311853	5.119001	6.910651	0.307692	0.324637	0.415385	0.389517	0.051282	0.346154	411.900066	591.065089	0.314998
5	6	0.100051	0.068038	2.694384	4.805224	0.161954	0.162521	0.288832	0.276165	0.134615	0.480769	169.438402	380.522366	0.405065
6	7	0.150013	0.053578	0.855360	3.489730	0.051414	0.059076	0.209760	0.203864	0.042735	0.523504	-14.463999	248.972969	0.397377
7	8	0.200103	0.048859	1.066458	2.883134	0.064103	0.050963	0.173299	0.165590	0.053419	0.576923	6.645847	188.313420	0.400919
8	9	0.300026	0.043784	0.812592	2.193544	0.048843	0.046085	0.131849	0.125789	0.081197	0.658120	-18.740799	119.354437	0.380995
9	10	0.400077	0.040092	0.726123	1.826571	0.043646	0.041845	0.109791	0.104796	0.072650	0.730769	-27.387731	82.657118	0.351841
10	11	0.500000	0.037186	0.684288	1.598291	0.041131	0.038627	0.096070	0.091572	0.068376	0.799145	-31.571199	59.829060	0.318276
11	12	0.600051	0.034513	0.726123	1.452867	0.043646	0.035848	0.087329	0.082281	0.072650	0.871795	-27.387731	45.286705	0.289122
12	13	0.699974	0.031589	0.384912	1.300414	0.023136	0.033054	0.078165	0.075254	0.038462	0.910256	-61.508800	30.041402	0.223730
13	14	0.800026	0.028206	0.320348	1.177847	0.019255	0.029974	0.070798	0.069591	0.032051	0.942308	-67.965176	17.784680	0.151381
14	15	0.899949	0.023525	0.320760	1.082683	0.019280	0.026037	0.065078	0.064755	0.032051	0.974359	-67.924000	8.268288	0.079169
15	16	1.000000	0.000692	0.256279	1.000000	0.015404	0.018306	0.060108	0.060108	0.025641	1.000000	-74.372140	0.000000	0.000000

		0	1	Error	Rate
0	0	1768.0	62.0	0.0339	(62.0/1830.0)
1	1	63.0	54.0	0.5385	(63.0/117.0)
2	Total	1831.0	116.0	0.0642	(125.0/1947.0)

	metric	threshold	value	idx
0	max f1	0.272361	0.463519	96.0
1	max f2	0.075372	0.509404	139.0
2	max f0point5	0.291926	0.467890	87.0
3	max accuracy	0.418666	0.941448	13.0
4	max precision	0.586764	1.000000	0.0
5	max recall	0.019987	1.000000	367.0
6	max specificity	0.586764	1.000000	0.0
7	max absolute_mcc	0.272361	0.429381	96.0
8	max min_per_class_accuracy	0.044083	0.726496	226.0
9	max mean_per_class_accuracy	0.053603	0.763311	184.0
10	max tns	0.586764	1830.000000	0.0
11	max fns	0.586764	116.000000	0.0
12	max fps	0.000947	1830.000000	399.0
13	max tps	0.019987	117.000000	367.0
14	max tnr	0.586764	1.000000	0.0
15	max fnr	0.586764	0.991453	0.0
16	max fpr	0.000947	1.000000	399.0
17	max tpr	0.019987	1.000000	367.0

	group	cumulative_data_fraction	lower_threshold	lift	cumulative_lift	response_rate	score	cumulative_response_rate	cumulative_score	capture_rate	cumulative_capture_rate	gain	cumulative_gain	kolmogorov_smirnov
0	1	0.010272	0.413237	8.320513	8.320513	0.500000	0.454803	0.500000	0.454803	0.085470	0.085470	732.051282	732.051282	0.080006
1	2	0.020031	0.378547	7.882591	8.107166	0.473684	0.395600	0.487179	0.425961	0.076923	0.162393	688.259109	710.716634	0.151464
2	3	0.030303	0.357108	4.992308	7.051282	0.300000	0.366802	0.423729	0.405907	0.051282	0.213675	399.230769	605.128205	0.195096
3	4	0.040062	0.335906	9.634278	7.680473	0.578947	0.345012	0.461538	0.391074	0.094017	0.307692	863.427800	668.047337	0.284741
4	5	0.050334	0.306114	9.152564	7.980900	0.550000	0.320887	0.479592	0.376750	0.094017	0.401709	815.256410	698.090005	0.373841
5	6	0.100154	0.065814	3.259582	5.632347	0.195876	0.149180	0.338462	0.263549	0.162393	0.564103	225.958234	463.234714	0.493611
6	7	0.149974	0.054028	1.544013	4.274236	0.092784	0.058728	0.256849	0.195509	0.076923	0.641026	54.401269	327.423604	0.522446
7	8	0.200308	0.048576	0.339613	3.285536	0.020408	0.051136	0.197436	0.159230	0.017094	0.658120	-66.038723	228.553583	0.487081
8	9	0.299949	0.043804	0.686228	2.422067	0.041237	0.045867	0.145548	0.121572	0.068376	0.726496	-31.377214	142.206709	0.453818
9	10	0.400103	0.040478	0.341354	1.901221	0.020513	0.042151	0.114249	0.101691	0.034188	0.760684	-65.864563	90.122116	0.383635
10	11	0.500257	0.037394	0.426693	1.606013	0.025641	0.038957	0.096509	0.089131	0.042735	0.803419	-57.330703	60.601274	0.322544
11	12	0.599897	0.034389	0.428892	1.410498	0.025773	0.035801	0.084760	0.080273	0.042735	0.846154	-57.110759	41.049789	0.262001
12	13	0.700051	0.031757	0.256016	1.245330	0.015385	0.033067	0.074835	0.073520	0.025641	0.871795	-74.398422	24.532987	0.182724
13	14	0.799692	0.028399	0.600449	1.164979	0.036082	0.030235	0.070006	0.068127	0.059829	0.931624	-39.955062	16.497867	0.140367
14	15	0.899846	0.023668	0.426693	1.082806	0.025641	0.026047	0.065068	0.063443	0.042735	0.974359	-57.330703	8.280646	0.079277
15	16	1.000000	0.000837	0.256016	1.000000	0.015385	0.018182	0.060092	0.058910	0.025641	1.000000	-74.398422	0.000000	0.000000

		0	1	Error	Rate
0	0	6980.0	338.0	0.0462	(338.0/7318.0)
1	1	266.0	202.0	0.5684	(266.0/468.0)
2	Total	7246.0	540.0	0.0776	(604.0/7786.0)

	metric	threshold	value	idx
0	max f1	0.118298	0.400794	189.0
1	max f2	0.062424	0.420387	231.0
2	max f0point5	0.305414	0.413286	117.0
3	max accuracy	0.420758	0.940663	44.0
4	max precision	0.825835	1.000000	0.0
5	max recall	0.016316	1.000000	384.0
6	max specificity	0.825835	1.000000	0.0
7	max absolute_mcc	0.221890	0.361377	159.0
8	max min_per_class_accuracy	0.043686	0.669445	280.0
9	max mean_per_class_accuracy	0.062424	0.701141	231.0
10	max tns	0.825835	7318.000000	0.0
11	max fns	0.825835	467.000000	0.0
12	max fps	0.000928	7318.000000	399.0
13	max tps	0.016316	468.000000	384.0
14	max tnr	0.825835	1.000000	0.0
15	max fnr	0.825835	0.997863	0.0
16	max fpr	0.000928	1.000000	399.0
17	max tpr	0.016316	1.000000	384.0

	group	cumulative_data_fraction	lower_threshold	lift	cumulative_lift	response_rate	score	cumulative_response_rate	cumulative_score	capture_rate	cumulative_capture_rate	gain	cumulative_gain	kolmogorov_smirnov
0	1	0.010018	0.420697	8.744959	8.744959	0.525641	0.502838	0.525641	0.502838	0.087607	0.087607	774.495946	774.495946	0.082551
1	2	0.020036	0.379848	7.251918	7.998439	0.435897	0.398132	0.480769	0.450485	0.072650	0.160256	625.191760	699.843853	0.149188
2	3	0.030054	0.348871	7.038626	7.678501	0.423077	0.364404	0.461538	0.421791	0.070513	0.230769	603.862590	667.850099	0.213551
3	4	0.040072	0.329732	6.398751	7.358563	0.384615	0.339837	0.442308	0.401303	0.064103	0.294872	539.875082	635.856345	0.271095
4	5	0.050090	0.300396	5.332292	6.953309	0.320513	0.315074	0.417949	0.384057	0.053419	0.348291	433.229235	595.330923	0.317271
5	6	0.100051	0.063827	2.523312	4.741154	0.151671	0.137477	0.284981	0.260925	0.126068	0.474359	152.331202	374.115401	0.398245
6	7	0.175700	0.060325	0.706144	3.003858	0.042445	0.060940	0.180556	0.174820	0.053419	0.527778	-29.385602	200.385802	0.374594
7	8	0.200103	0.054462	0.437809	2.690925	0.026316	0.057170	0.161746	0.160473	0.010684	0.538462	-56.219073	169.092525	0.359997
8	9	0.300026	0.045913	1.005048	2.129447	0.060411	0.049340	0.127997	0.123460	0.100427	0.638889	0.504801	112.944730	0.360534
9	10	0.400077	0.041583	0.683410	1.767822	0.041078	0.043692	0.106260	0.103512	0.068376	0.707265	-31.659041	76.782182	0.326833
10	11	0.500000	0.038204	0.620136	1.538462	0.037275	0.039807	0.092474	0.090781	0.061966	0.769231	-37.986399	53.846154	0.286449
11	12	0.600051	0.035295	0.726123	1.403014	0.043646	0.036731	0.084332	0.081769	0.072650	0.841880	-27.387731	40.301377	0.257294
12	13	0.699974	0.032230	0.598752	1.288204	0.035990	0.033765	0.077431	0.074916	0.059829	0.901709	-40.124800	28.820356	0.214636
13	14	0.800026	0.028855	0.320348	1.167163	0.019255	0.030608	0.070156	0.069375	0.032051	0.933761	-67.965176	16.716338	0.142288
14	15	0.899949	0.023951	0.363528	1.077934	0.021851	0.026594	0.064792	0.064625	0.036325	0.970085	-63.647200	7.793428	0.074622
15	16	1.000000	0.000326	0.298992	1.000000	0.017972	0.018671	0.060108	0.060027	0.029915	1.000000	-70.100831	0.000000	0.000000

		mean	sd	cv_1_valid	cv_2_valid	cv_3_valid	cv_4_valid	cv_5_valid	cv_6_valid	cv_7_valid	cv_8_valid	cv_9_valid	cv_10_valid	cv_11_valid	cv_12_valid	cv_13_valid	cv_14_valid	cv_15_valid	cv_16_valid	cv_17_valid	cv_18_valid	cv_19_valid	cv_20_valid	cv_21_valid	cv_22_valid	cv_23_valid	cv_24_valid	cv_25_valid	cv_26_valid	cv_27_valid	cv_28_valid	cv_29_valid	cv_30_valid
0	accuracy	0.9293634	0.025219189	0.95	0.9230769	0.9269231	0.95384616	0.95384616	0.9230769	0.96153843	0.9153846	0.85384613	0.9153846	0.93846154	0.9346154	0.9269231	0.88846153	0.9115385	0.9692308	0.8996139	0.9189189	0.9189189	0.9459459	0.93436295	0.9150579	0.9227799	0.96525097	0.9498069	0.8918919	0.9459459	0.93050194	0.95752895	0.93822396
1	auc	0.76348716	0.071166314	0.76800233	0.71019876	0.79448074	0.85323113	0.76584023	0.73025554	0.6888441	0.7904345	0.7159797	0.77731866	0.736168	0.8148148	0.54239255	0.7096774	0.7122175	0.8783602	0.78377193	0.73764855	0.7493421	0.8430269	0.82401556	0.7942464	0.71366066	0.8042877	0.85510206	0.832636	0.77339786	0.7363683	0.8395722	0.6293223
2	err	0.07063657	0.025219189	0.05	0.07692308	0.073076926	0.046153847	0.046153847	0.07692308	0.03846154	0.08461539	0.14615385	0.08461539	0.06153846	0.06538462	0.073076926	0.11153846	0.08846154	0.03076923	0.1003861	0.08108108	0.08108108	0.054054055	0.06563707	0.08494209	0.077220075	0.034749035	0.05019305	0.10810811	0.054054055	0.06949807	0.042471044	0.06177606
3	err_count	18.333334	6.5513005	13.0	20.0	19.0	12.0	12.0	20.0	10.0	22.0	38.0	22.0	16.0	17.0	19.0	29.0	23.0	8.0	26.0	21.0	21.0	14.0	17.0	22.0	20.0	9.0	13.0	28.0	14.0	18.0	11.0	16.0
4	f0point5	0.45547372	0.15147571	0.530303	0.4225352	0.4494382	0.42553192	0.6818182	0.32051283	0.5681818	0.29126215	0.12269939	0.4385965	0.4861111	0.505618	0.3030303	0.20833333	0.5092593	0.6818182	0.36036035	0.29411766	0.43956044	0.49382716	0.46153846	0.30927834	0.36231884	0.8064516	0.530303	0.36290324	0.6034483	0.48076922	0.71428573	0.5
5	f1	0.43831208	0.11235892	0.5185185	0.375	0.45714286	0.4	0.6	0.33333334	0.5	0.3529412	0.17391305	0.47619048	0.46666667	0.51428574	0.2962963	0.25641027	0.4888889	0.6	0.3809524	0.32258064	0.43243244	0.53333336	0.41379312	0.3529412	0.33333334	0.6896552	0.5185185	0.39130434	0.5	0.5263158	0.56	0.3846154
6	f2	0.43535635	0.089564964	0.5072464	0.33707866	0.4651163	0.3773585	0.53571427	0.3472222	0.44642857	0.4477612	0.29850745	0.5208333	0.44871795	0.5232558	0.28985506	0.33333334	0.47008547	0.53571427	0.4040404	0.35714287	0.42553192	0.5797101	0.375	0.41095892	0.30864197	0.60240966	0.5072464	0.4245283	0.42682928	0.5813953	0.46052632	0.3125
7	lift_top_group	8.089831	5.756644	12.380953	9.122807	5.098039	15.757576	14.444445	0.0	7.2222223	0.0	0.0	9.62963	10.833333	5.098039	0.0	0.0	7.2222223	21.666666	13.631579	6.6410255	4.5438595	6.6410255	0.0	13.282051	5.0784316	14.388889	6.1666665	8.633333	14.388889	10.791667	15.235294	4.796296
8	logloss	0.18792725	0.032492485	0.16183199	0.23102114	0.20027487	0.13980255	0.19036962	0.18677619	0.15548834	0.15646584	0.1322693	0.20070384	0.19327386	0.19256084	0.20043705	0.17882729	0.26566097	0.13172741	0.22183247	0.18320592	0.22141643	0.14802365	0.20704313	0.1712321	0.22069813	0.17449671	0.15825497	0.22827521	0.20340662	0.17768437	0.17275468	0.23200172
9	max_per_class_error	0.55388075	0.09442896	0.5	0.68421054	0.5294118	0.6363636	0.5	0.64285713	0.5833333	0.45454547	0.42857143	0.44444445	0.5625	0.47058824	0.71428573	0.5833333	0.5416667	0.5	0.57894737	0.61538464	0.57894737	0.3846154	0.64705884	0.53846157	0.7058824	0.44444445	0.5	0.55	0.6111111	0.375	0.5882353	0.7222222
10	mcc	0.4140981	0.11867936	0.49256343	0.3424201	0.41819814	0.37829927	0.58992136	0.29344437	0.49154666	0.33825287	0.19631688	0.43646932	0.43525764	0.47948983	0.25800943	0.22556594	0.44188073	0.59769464	0.3286362	0.28485996	0.38897252	0.5103111	0.3866006	0.32041422	0.29609627	0.6953352	0.49245626	0.33654246	0.49686882	0.497063	0.5834707	0.3899857
11	mean_per_class_accuracy	0.70338625	0.04551903	0.7378049	0.64337194	0.714718	0.671778	0.74380165	0.6562137	0.70228493	0.7385907	0.71654433	0.74885213	0.7044057	0.7461874	0.62456447	0.6639785	0.7079802	0.74596775	0.6792763	0.6658849	0.689693	0.7893996	0.6640739	0.70028144	0.6305299	0.7757031	0.7377551	0.6894351	0.6882204	0.78780866	0.70381624	0.6326648
12	mean_per_class_error	0.29661378	0.04551903	0.2621951	0.3566281	0.28528202	0.32822198	0.25619835	0.3437863	0.29771507	0.26140928	0.28345567	0.25114784	0.29559427	0.25381264	0.37543553	0.3360215	0.29201978	0.25403225	0.32072368	0.33411506	0.31030703	0.21060038	0.33592612	0.2997186	0.3694701	0.22429691	0.2622449	0.31056485	0.31177962	0.21219136	0.29618376	0.36733517
13	mse	0.048360467	0.009946226	0.041215982	0.05927394	0.053386357	0.03490668	0.04873468	0.04785549	0.03744534	0.039178222	0.027377248	0.05329377	0.049598545	0.051712077	0.049356885	0.043730482	0.0722259	0.031440783	0.059075277	0.04569484	0.059118953	0.038147986	0.0559402	0.043693185	0.057063483	0.044180162	0.041129515	0.06255136	0.052951276	0.045717184	0.04519618	0.05962203
14	null_deviance	118.02124	19.058928	109.23703	136.7752	125.73113	92.82863	131.24835	109.23703	98.28862	92.82863	71.082695	131.24835	120.223526	125.73113	109.23703	98.28862	164.55519	98.28862	136.65318	103.63261	136.65318	103.63261	125.607956	103.63261	125.607956	131.12575	109.11213	142.19029	131.12575	120.09978	125.607956	131.12575
15	pr_auc	0.30679077	0.13954958	0.29536867	0.32186282	0.28712222	0.36642668	0.5025384	0.15924418	0.25598085	0.1574329	0.06026311	0.37355992	0.31441343	0.31143665	0.10196809	0.102661744	0.3163861	0.5751905	0.3811229	0.15434223	0.26492032	0.3232288	0.26717553	0.24867943	0.17561266	0.6126563	0.33455437	0.34606618	0.45739362	0.31824458	0.58488417	0.23298557
16	precision	0.47607982	0.1905555	0.53846157	0.46153846	0.44444445	0.44444445	0.75	0.3125	0.625	0.26086956	0.102564104	0.41666666	0.5	0.5	0.30769232	0.18518518	0.52380955	0.75	0.3478261	0.2777778	0.44444445	0.47058824	0.5	0.2857143	0.3846154	0.90909094	0.53846157	0.34615386	0.7	0.45454547	0.875	0.625
17	r2	0.13546097	0.08348212	0.19099875	0.12493597	0.12638155	0.13848421	0.24369505	0.060676213	0.14942709	0.033060268	-0.045003917	0.17294337	0.1411727	0.15377963	0.031206282	0.006659745	0.13798183	0.2858209	0.1309586	0.041508563	0.13031615	0.19981082	0.087864734	0.08349518	0.069549	0.3168166	0.1956242	0.122174166	0.18118382	0.21122575	0.2630518	0.0780301
18	recall	0.44611925	0.09442896	0.5	0.31578946	0.47058824	0.36363637	0.5	0.35714287	0.41666666	0.54545456	0.5714286	0.5555556	0.4375	0.5294118	0.2857143	0.41666666	0.45833334	0.5	0.42105263	0.3846154	0.42105263	0.61538464	0.3529412	0.46153846	0.29411766	0.5555556	0.5	0.45	0.3888889	0.625	0.4117647	0.2777778
19	residual_deviance	97.54081	16.838604	84.15263	120.13099	104.14293	72.69732	98.9922	97.12362	80.853935	81.36224	68.78003	104.366	100.5024	100.13164	104.22727	92.99019	138.1437	68.49825	114.90922	94.900665	114.6937	76.67625	107.24834	88.69823	114.32163	90.3893	81.976074	118.24656	105.364624	92.0405	89.48692	120.17689